>A-450/3-80-033a
VOC Fugitive Emissions
in Synthetic Organic Chemicals
Manufacturing 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
November 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 Officer (MD-35),
U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, or from the National Technical
Information Services, 5285 Port Royal Road, Springfield,
Virginia 22161.
Publication No. EPA-450/3-80-033a
n
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-450/3-80-033a
2.
fugitive7""Emissions in Synthetic Organic Chemicals
Manufacturing Industry - Background Information for
Proposed Standards
7. AUTHOR(S)
9 PERFORMING OBGANIZATJQN NAME AND. ADDRESS
OffTee of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
i c 2> 1 £. y
I j s. 4. Q y
5. REPORT DATE '
November 1980 ':_•_ __.
6. PERFORMING ORGANIZATION CODh
OHT NO
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3058
12.
iNSDRING^GE.WCy-NAME AMD ADDRESS > r-j. J J
for Air T)uality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OH REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200704
15. SUPPLEMENTARY NOTES
16. ABSTRACT
v-JStandards of performance to control fugitive emissions of VOC from new, modified,
and reconstructed Synthetic Organic Chemical Manufacturing Industry (SOCMI) plants
are beinq proposed under Section 111 of the Clean Air Act. This document contains
information on SOCMI, emission control technology for fugitive emissions of VOC,
Regulatory Alternatives which were considered, analyses of environmental, energy,
costs, and other technical data to support the standard of performance.^
KEV WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air pollution
Pollution Control
Standards of performance
Volatile Organic Compounds
Organic Chemical Industry
b.lDC NTIFIERS/OPEN fcNDtD TERMS C. COSATI I'ield/Group
Air pollution control
B. Olf TP!OUTION
. . ^_ --- ^ — -- ---- - — -- ;r^-r
\ Re 1 ea se_ jj nrurn tecL_ ^.Ay a^ajjj e ^ f iioni .EPA •.,_. '
fc "ribrVr7'^MJ)-^||;:,l\Re"se|,rqh' Jriangl e'^Rark '
!fv North 1:fr¥Vf na^y-yTT-"^" ^' "T ll^l :1-Z :' l
19. SECUPITY CLASS (This Report/
unclassified
2O SECURITY CLAfiS (This
unclassified
21. NO. OF PAGES
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION IS
-------
-------
/y
ENVIRONMENTAL PROTECTION AGENCY
Background Information and Draft
Environmental Impact Statement for
VOC Fugitive Emissions in Synthetic Organic
Chemicals Manufacturing Industry
Prepared by:
9-31
Don R. Goodwill (Date)
Director, Emission Standards and Engineering Division
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
1. The proposed standards of performance would limit emissions of
volatile organic compounds from new, modified, and reconstructed
units in the synthetic organic chemicals manufacturing industry.
Section 111 of the Clean Air Act (42 U.S.C. 7411), as amended,
directs the Administrator to establish standards of performance for
any category of new stationary source of air pollution that ". . .
causes or contributes significantly to air pollution which may
reasonably be anticipated to endanger public health or welfare."
2. Copies of this document have been sent to the following: Federal
Departments of Labor, Health and Human Services, Defense, Transportation,
Agriculture, Commerce, Interior, and Energy; the National Science
Foundation; the Council on Environmental Quality; members of the
State and Territorial Air Pollution Program Administrators; the
Association of Local Air Pollution Control Officials; EPA Regional
Administrators; and other interested parties.
3. The comment period for review of this document is 75 days and is
expected to begin on or about December 15.
4. For additional information contact:
Ms. Susan Wyatt
Standards Development Branch (MD-13)
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
telephone: (919) 541-5477
5. Copies of this document may be obtained from:
U. S. EPA Library (MD-35)
Research Triangle .Park, .NC .27711 . ..
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
iii
"i
Preceding page blank
-------
METRIC CONVERSION TABLE
In keeping with U.S. Environmental Protection Agency policy, metric
units are used in this report. These units may be converted to common
English units by using the following conversion factors:
Equivalent
Metric Unit Metric Name English Unit
LENGTH
m meter 39.3700 in.
m meter 3.2810 ft.
VOLUME
13 liters 0.2642 U.S. gal
m3 cubic meters 264.2 U.S. gal
m cubic meters 6.29 Barrels (bbl)
WEIGHT
Kg kilogram (10* grams) 2.2046 Ib.
Mg megagram (10g grams) 1.1023 tons
Gg gigagram (10 grams) 1,102.3 tons
ENERGY
GJ giga.joule 9.48 X 10'' Btu
GJ gigajoule 277.76 KWh
J/g joule per gram 0.430 Btu/lb.
VOLUMETRIC FLOW
Nm3/sec normal cubic meters per second 2242 SCFM (ft /min)
SPEED
m/s meters per second 196.86 ft/min
Temperature in degrees Celcius (°C) can be converted to temperature
in degrees Farenheit (°F) by the following formula:
(°F) = 1.8 (°C) + 32
IV
-------
TABLE OF CONTENTS
METRIC CONVERSION TABLE . . -jv
TABLE OF CONTENTS . v
LIST OF TABLES ..•••' vii
LIST OF FIGURES . . xi
1. SUMMARY 1-1
1.1 Regulatory Alternatives 1-1
1.2 Environmental Impact . 1-2
1.3 Economic Impact 1-3
2. INTRODUCTION 2-1
2.1 Background and Authority for Standards 2-1
2.2 Selection of Categories of Stationary Sources 2-5
2.3 Procedure for Development of Standards of
Performance 2-7
2.4 Consideration of Costs 2-9
2.5 Consideration of Environmental Impacts 2-10
2.6 Impact on Existing Sources 2-11
2.7 Revision of Standards of Performance 2-12
3. DESCRIPTION OF FUGITIVE EMISSION SOURCES 3-1
3.1 Introduction and General Industry Information 3-1
3.2 Fugitive Emission Definition and Potential 3-3
3.3 Baseline Control 3-17
3.4 References 3-22
4. EMISSION CONTROL TECHNIQUES 4-1
4.1 Leak Detection and Repair Methods 4-1
4.2 Equipment Specifications 4-13
4.3 References 4-25
5. MODIFICATION AND RECONSTRUCTION 5-1
5.1 General Discussion of Modification
and Reconstruction 5-1
5.2 Applicability of Modification and Reconstruction
Provisions' ''to the" SOCMI 5-3
-------
TABLE OF CONTENTS (cont.)
Page
6. MODEL PROCESS UNITS AND REGULATORY ALTERNATIVES ... 6-1
6.1 Model Units 6-2
6.2 Regulatory Alternatives 6-4
6.3 References 6-7
7. ENVIRONMENTAL IMPACT 7-1
7.1 Impact on Atmospheric Emissions 7-1
7.2 Impact of Water Quality 7-8
7.3 Impact on Solid Waste 7-11
7.4 Energy Impact 7-12
7.5 Other Environmental Concerns 7-14
7.6 References 7-14
8. COST ANALYSIS 8-1
8.1 Cost Analysis of Regulatory Alternatives 8-1
8.2 Other Cost Considerations . 8-21
8.3 References 8-28
9. ECONOMIC ANALYSIS 9-1
9.1 Industry Profile 9-1
9.2 Economic Impact Analysis : 9-16
9.3 Socio-Economic and Inflation.iry Impacts . 9-35
9.4 References 9-36
APPENDIX A A_T
APPENDIX B ...'.'...... B-l
APPENDIX C C-1
APPENDIX D D-i
APPENDIX E E_-|
APPENDIX F . . F-l
APPENDIX G G-l
VI
-------
LIST OF TABLES
Page
1-1 Environmental and Economic Impacts of Regulatory Alternatives . . . 1-4
3-1 Uncontrolled Fugitive Emission Factors in the Synthetic Organic
Chemical Manufacturing Industry (SOCMI) 3-19
4-1 Fraction of Total Mass Emissions From Various Source Types That
Would be Affected by Different Action Levels 4-7
4-2 Estimated Occurrence and Recurrence Rate of Leaks for Various
Monitoring Intervals 4-10
4-3 Maximum Potential Control Efficiency as a Function of Repair
Interval Assuming 100 Percent Efficiency for Other Factors 4-11
4-4 Average Emission Rates From Sources Above 10,000 PPMV and at
1000 PPMV 4-11
4-5 Example of Control Efficiency Calculation 4-14
4-6 Impact of Monitoring Interval on Correction Factor Accounting for
Leak Occurrence/Recurrence (for Example Calculation) 4-15
4-7 Effectiveness of Equipment Modifications 4-24
6-1 Fugitive Emission Sources for Three Model Units 6-3
6-2 Regulatory Alternatives for Fugitive Emission Sources in SOCMI. . . 6-5
7-1 Emission Factors for Sources Controlled Under Regulatory
Alternative II 7-3
7-2 Emission Factors for Sources Controlled Under Regulatory
Alternative III 7-4
7-3 Emission Factors for Sources Controlled Under Regulatory
Alternative IV 7-5
7-4 Example Calculation of VOC Fugitive Emissions From Model Unit A
Under Regulatory Alternative IT . 7-6
7-5 Estimated Emissions and Emission Reductions on a Model
Unit Basis 7-7
7-6 Total VOC Fugitive Emissions From Affected Model Units for
Regulatory Alternatives- . 7-10
Vll
-------
7-7 Energy Impact of Emission Reductions for Regulatory Alternatives. . 7-13
8-1 Capital Cost Data 8-2
8-2 Capital Cost Estimates for New Model Units 8-4
8-3 Annual Monitoring and Leak Repair Labor Requirements for
Regulatory Alternative II 8-6
8-4 Annual Monitoring and Leak Repair Labor Requirements for
Regulatory Alternative III 8-7
8-5 Annual Monitoring and Leak Repair Labor Requirements for
Regulatory Alternative IV 8-8
8-6 Derivation of Annualized Labor, Administrative, Maintenance and
Capital Charges 8-10
8-7 Labor-Hour Requirements for Initial Leak Repair 8-12
8-8 Recovery Credits 8-13
8-9 Annualized Control Cost Estimates for Model Unit A 8-14
8-10 Annualized Control Cost Estimates for Model Unit B 8-15
8-11 Annualized Control Cost Estimates for Model Unit C 8-16
8-12 Cost Effectiveness for Model Units 8-17
8-13 Capital Cost Estimates for Modified/Reconstructed Facilities. . . . 8-19
8-14 Annualized Control Cost Estimates for Modified/Reconstructed
Model Units Under Regulatory Alternative IV 8-20
8-15 Nationwide Costs for the Industry Under Regulatory
Alternative II 8-22
8-16 Nationwide Costs for the Industry Under Regulatory
Alternative III 8-23
8-17 Nationwide Costs for the Industry Under Regulatory
Alternative IV 8-24
8-18 Statutes That May be Applicable to SOCMI 8-25
9-1 Estimated Annual Production Capacity by State, 1976 9-3
9-2 Distribution of Units by Unit Capacity and Region, 1976 9-4
vi n
-------
9-3 Distribution of Industry Capacity by Unit Capacity and ' - (
Region, 1976 9-5
9-4 Annual Production and Sales of Synthetic Organic Chemicals 9-7
9-5 SOCMI Resource Use 9-8
9-6 Industrial Organic Chemicals: U.S. Imports and Exports, 1966-77. . 9-11
9-7 Industrial Organic Chemicals: U.S. Trade, by Principal Trading
Partners, 1976 and 1977 9-12
9-8 Industrial Organic Chemicals: U.S. Imports for Consumption,
by Principal Sources, 1972-77 9-13
9-9 Industry Concentration, 1976 9-15
9-10 Estimated Cost of Capital for Firms in SOCMI 9-18
9-11 Average Rate of Return Impacts 9-25
9-12 Model Units Experiencing Significant Rate of Return Impacts
Under Full Cost Absorption 9-27
9-13 Average Percentage Price Impacts of Regulatory Alternatives .... 9-27
9-14 Model Units Requiring Significant Price Increases to Maintain
Target Rates of Return 9-29
9-15 Investment Impacts 9-31
9-16 Employment Impacts 9-33
9-17 Model Unit and Industry Annualized Control Costs 9-34
C-l Frequency of Leaks From Fugitive Emission Sources in Synthetic
Organic Chemical Units C-4
C-2 Twenty-four Chemical Process Units Screened for
Fugitive Emissions C-6
C-3 Summary of SOCMI Process Units Fugitive Emissions C-8
C-4 Average Fugitive Emission Source Screening Rates C-9
C-5 Sampled Process Units From Nine Refineries During Refinery Study. . C-10
C-6 Leak Frequencies and Emission Factors From Fugitive Sources in
Petroleum Refineries C-12
IX
-------
C-7 Comparison of Leak Frequencies for Fugitive Emission Sources in
SOCMI Units and Petroleum Refineries C-13
C-8 Frequency of Leaks From Fugitive Emission Sources in
Two DuPont Plants . . . C-15
C-9 Frequency of Leaks From Fugitive Emission Sources in
Exxon's Cyclohexane Unit C-16
C-10 Summary of Maintenance Study Results From the Union Oil Co.
Refinery in Rodeo, California C-18
C-ll Summary of Maintenance Study Results From the Shell Oil Company
Refinery in Martinez, California C-20
C-12 Summary of EPA Refinery Maintenance Study Results C-22
C-13 Unit D Ethylene Unit Block Valve Repairs C-23
E-l Yields by Rating Class for Cost of Debt Funds, 1979 E-5
E-2 Financial Data for 100 Chemical Firms E-7
G-l Uncontrolled Emissions Estimates From the Model Units G-l
-------
LIST OF FIGURES
3-1 General schematic of process levels that make up the organic
chemical industry 3-2
3-2 Diagram of a simple packed seal 3-4
3-3 Diagram of a basic single mechanical seal 3-5
3-4 Diagram of a dual mechanical seal 3-6
3-5 Diagram of a dual mechanical seal 3-6
3-6 Diaphragm Pump 3-7
3-7 Labyrinth compressor seal 3-9
3-8 Restrictive ring compressor seal 3-9
3-9 Mechanical contact compressor seal 3-10
3-10 Liquid film compressor seal 3-10
3-11 Diagram of a gate valve 3-11
3-12 Example of bellows seals 3-12
3-13 Diagrams of valves with diaphragm seals 3-13
3-14 Diagram of a spring-loaded relief valve 3-14
3-15 Diagram of hydraulic seal for agitators 3-16
3-16 Diagram of agitator lip seal 3-16
4-1 Cumulative distribution of total emissions by screening values -
valves - gas/vapor streams 4-16
4-2 Cumulative distribution of sources by screening values - valves -
gas/vapor streams 4-16
-------
-------
1.0 SUMMARY
1.1 REGULATORY ALTERNATIVES
Standards of performance for new stationary sources of volatile
organic compounds (VOC) from fugitive emission sources in the Synthetic
Organic Chemicals Manufacturing Industry (SOCMI) are being developed
under the authority of Section 111 of the Clean Air Act. These standards
would affect new stationary sources which produce as final products or
intermediates one or more of certain organic chemicals. These standards
would reduce emissions from pumps, compressors, valves, safety/relief
valves, sampling connections, and open-ended lines.
Four regulatory alternatives were considered. Regulatory Alternative I
is the baseline alternative and represents the level of control that
would exist in the absence of any standards of performance. Requirements
of Alternative II corresponds to the requirements of the Control Techniques
Guidelines document (EPA-450/2-78-036) for petroleum refineries. These
requirements are:
• Quarterly monitoring of all in-line valves, open-ended valves
and safety/relief valves in gas service (relief valves would
also be monitored after overpressure relieving to check for
proper reseating);
• Annual monitoring of all in-line valves and open-ended valves
in light liquid service;
• Quarterly monitoring of compressor seals;
• Annual monitoring of light liquid service pumps (such pumps would
also be inspected visually for liquid leaks each week; immediate
instrument monitoring of visually leaking pumps would be required);
and
• Installation of caps, blinds, plugs, or second valves to seal
all open-ended lines.
1-1
-------
Regulatory Alternative III is more restrictive than Alternative II.
Monthly rather than quarterly or annual monitoring would be required.
Also, caps, plugs, or second valves would be required on open-ended
lines, and weekly pump inspections as for Alternative II.
Of the four alternatives, Regulatory Alternative IV is the most
restrictive. The requirements are:
•Monthly monitoring of all in-line valves and open-ended valves
in gas and light liquid service;
•Installation of rupture disks upstream of gas service safety/relief
valves that vent to the atmosphere (the disk would be replaced
if disk failure were detected);
•Installation of closed vents and control devices for compressor
seal area and/or degassing vents from compressor seal oil reservoirs;
•Installation of double mechanical seals on pumps in light liquid
service and installation of closed vent control devices for degassing
vents from seal oil reservoirs of all pumps in light liquid service
(weekly visual inspections of pumps in light liquid service would
also be required, with subsequent instrument monitoring required
for those pumps with visible liquid leaks);
• Installation of closed loop sampling systems; and
• Installation of caps, blinds, plugs, or second valves to seal
all open-ended lines.
1.2 ENVIRONMENTAL IMPACT
Fugitive emissions of VOC from affected SOCMI facilities would be
200 Gg/yr under Alternative I compared to 73, 62, and 26 Gg/yr under
Alternatives II, III, and IV. Emissions reductions effected by Alternatives II,
III, and IV would be 63, 69, and 87 percent, respectively.
In addition to reducing emission to the atmosphere, Alternatives II
and III would reduce liquid leaks which might otherwise become a part of
wastewater streams. Reduction of pollutants in effluents would also
reduce wastewater treatment needs. Implementation of Alternative IV
would also reduce liquid leaks, thereby reducing wastewater treatment
needs. However, a small amount of washewater containing suspened solids
and some solid waste could result from the use of control systems required
1-2
-------
by Alternative IV. The impact of the wastewater and solid waste would
be very smal1.
Energy savings would result under Regulatory Alternatives II, III,
and IV. Under Alternative II, VOC's recovered during the fifth year of
implementation would have an energy content of about 3,940 TJ. This heating
value is equivalent to the heating value of 644,000 barrels of crude oil.
VOC recovered under Alternative III in the fifth year would have a heating
value of about 4,250 TJ which is equivalent to the heating value of 695,000
barrels of crude. The heating value of VOC recovered under Alternative IV
would be 5,360 TJ. This is the same heating value found in 876,000 barrels
of crude oil.
A more detailed analysis of environmental and energy impacts is presented
in Chapter 7. A summary of the environmental and economic impacts associated
with the four regulatory alternatives is shown in Table 1-1.
1.3 ECONOMIC IMPACT
Costs incurred by SOCMI under Regulatory Alternatives II and III would
actually be credits due to the value of recovered VOC. In the fifth year
after implementation of Alternative II, a net annualized credit of $29 million
would result. For the same year under Alternative III, a net annualized
credit of $21 million would result. Net annualized costs incurred during the
fifth year under Regulatory Alternative IV would be $11 million. In this
Alternative the costs exceed the value of recovered VOC. A more detailed
analysis of costs is included in Chapter 8.
In general, most units will not increase product prices as a result of
the implementation of Regulatory Alternatives II, III, or IV. A more
detailed economic analysis is presented in Chapter 9.
Economic impacts associated with the four Regulatory Alternatives are
shown in Table 1-1.
1-3
-------
TABLE 1-1. ENVIRONMENTAL AND ECONOMIC IMPACTS OF REGULATORY ALTERNATIVES
Administrative
Action
Regulatory
Alternative I
(No Action)
Regulatory
Alternative II
Regulatory
Alternative III
Regulatory
Alternative IV
Air
Impact
0
+2**
+2**
+3**
Water
Impact
0
+]**
+1**
+1**
Solid
Waste
Impact
0
0
0
0
Energy
Impact
0
+1*
+1*
+1*
Noise
Impact
0
0
0
0
Economic
Impact
0
+1*
+1*
-1*
KEY: + Beneficial impact
- Adverse impact
0 No impact
1 negligible impact
2 Small impact
3 Moderate impact
4 Large impact
* Short-term impact
** Long-term impact
*** Irreversible impact
-------
2. INTRODUCTION
2.1 BACKGROUND AND AUTHORITY FOR STANDARDS
Before standards of performance are proposed as a Federal regulation,
air pollution control methods available to the affected industry and the
associated costs of installing and maintaining the control equipment are
examined in detail. Various levels of control based on different technolo-
gies and degrees of efficiency are expressed as regulatory alternatives.
Each of these alternatives is studied by EPA as a prospective basis for
a standard. The alternatives are investigated in terms of their impacts
on the economics and well-being of the industry, the impacts on the
national economy, and the impacts on the environment. This document
summarizes the information obtained through these studies so that interested
persons will be able to see the information considered by EPA in the
development of the proposed standard.
Standards of performance for new stationary sources are established
under Section 111 of the Clean Air Act (42 ll.S.C. 7411) as amended,
hereinafter referred to as the Act. Section 111 directs the Administrator
to establish standards of performance for any category of new stationary
source of air pollution which ". . . causes, or contributes significantly
to air pollution which may reasonably be anticipated to endanger public
health or welfare."
The Act requires that standards of performance for stationary
sources reflect ". . . the degree of emission reduction achievable which
(taking into consideration the cost of achieving such emission reduction,
and any nonair quality health and environmental impact and energy require-
ments) the Administrator determines has been adequately demonstrated for
that category of sources." The standards apply only to stationary
sources, the construction or modification of which commences after
regulations are proposed by publication in the Federal Register.
-------
The 1977 amendments to the Act altered or added numerous provisions
that apply to the process of establishing standards of performance.
1. EPA is required to list the categories of major stationary sources
that have not already been listed and regulated under standards of perform-
ance. Regulations must be promulgated for these new categories on the
following schedule:
a. 25 percent of the listed categories by August 7, 1980.
b. 75 percent of the listed categories by August 7, 1981.
c. 100 percent of the listed categories by August 7, 1982.
A governor of a State may apply to the Administrator to add a category not
on the list or may apply to the Administrator to have a standard of perform-
ance revised.
2. EPA is required to review the standards of performance every 4
years and, if appropriate, revise them.
3. EPA is authorized to promulgate a standard based on design, equip-
ment, work practice, or operational procedures when a standard based on
emission levels is not feasible.
4. The term "standards of performance" is redefined, and a new term
"technological system of continuous emission reduction" is defined. The new
definitions clarify that the control system must be continuous and may
include a low- or non-polluting process or operation.
5. The time between the proposal and promulgation of a standard under
Section 111 of the Act may be extended to 6 months.
Standards of performance, by themselves, do not guarantee protection
of health or welfare because they are not designed to achieve any specific
air quality levels. Rather, they are designed to reflect the degree of
emission limitation achievable through application of the best adequately
demonstrated technological system of continuous emission reduction, taking
into consideration the cost of achieving such emission reduction, any
non-air-quality health and environmental ,impacts, and energy requirements.
Congress had several reasons for including these requirements. First,
standards with a degree of uniformity are needed to avoid situations
where some states may attract industries by relaxing standards relative to
other states. Second, stringent standards enhance the potential for
long-term growth. Third, stringent standards may help achieve long-term '
i
i
I
2-2
-------
cost savings by avoiding the need for more expensive retrofitting when
pollution ceilings may be reduced in the future. Fourth, certain types
of standards for coal-burning sources can adversely affect the coal
market by driving up the price of low-sulfur coal or effectively excluding
certain coals from the reserve base because their untreated pollution
potentials are high. Congress does not intend that new source performance
standards contribute to these problems. Fifth, the standard-setting
process should create incentives for improved technology.
Promulgation of standards of performance does not prevent State or
local agencies from adopting more stringent emission limitations for the
same sources. States are free under Section 116 of the Act to establish
even more stringent emission limits than those established under Section
111 or those necessary to attain or maintain the National Ambient Air
Quality Standards (NAAQS) under Section 110. Thus, new sources may in
some cases be subject to limitations more stringent than standards of
performance under Section 111, and prospective owners and operators of
new sources should be aware of this possibility in planning for such
facilities.
A similar situation may arise when a major emitting facility is to
be constructed in a geographic area that falls under the prevention of
significant deterioration of air quality provisions of Part C of the
Act. These provisions require, among other things, that major emitting
facilities to be constructed in such areas are to be subject to best
available control technology. The term Best Available Control Technology
(BACT), as defined in the Act, means
". . . an emission limitation based on the maximum degree of
reduction of each pollutant subject to regulation under this
Act emitted from, or which results from, any major emitting
facility, which the permitting authority, on a case-by-case
basis, taking into account energy, environmental, and economic
impacts and other costs, determines is achievable for such
facility through application of production processes and
available methods, systems, and techniques, including fuel
cleaning or treatment or innovative fuel combustion techniques
2-3
-------
for control of each such pollutant. In no event shall applica-
tion of "best available control technology" result in emissions
of any pollutants which will exceed the emissions allowed by
any applicable standard established pursuant to sections 111
or 112 of this Act. (Section 169(3))
Although standards of performance are normally structured in terms
of numerical emission limits where feasible, alternative approaches are
sometimes necessary. In some cases physical measurement of emissions
from a new source may be impractical or exorbitantly expensive. Section
lll(h) provides that the Administrator may promulgate a design or equipment
standard in those cases where it is not feasible to prescribe or enforce
a standard of performance. For example, emissions of hydrocarbons from
storage vessels for petroleum liquids are greatest during tank filling.
The nature.of the emissions, high concentrations for short periods
during filling and low concentrations for longer periods during storage,
and the configuration of storage tanks make direct emission measurement
impractical. Therefore, a more practical approach to standards of
performance for storage vessels has been equipment specification.
In addition, section lll(i) authorizes the Administrator to grant
waivers of compliance to permit a source to use innovative continuous
emission control technology. In order to grant the waiver, the
Administrator must find: (1) a substantial likelihood that the technology
will produce greater emission reductions than the standards require or
an equivalent reduction at lower economic, energy., of eaviropiuental cost;
(2) the proposed system has not been adequately demonstrated; (3) the
technology will not cause or contribute to an unreasonable risk to the
public health, welfare, or safety; (4) the governor of the State where
the source is located consents; and (5) the waiver will not prevent the
attainment or maintenance of any ambient standard. A waiver may have
conditions attached to assure the source will not prevent attainment of
any NAAQS. Any such condition will have the force of a performance
standard. Finally, waivers have definite end dates and may be terminated
earlier if the conditions are not mef or if the system fails to perform
as expected. In such a case, the source may he rjiven up to 3 years to
2-4
-------
to meet the standards with a mandatory progress schedule.
2.2 SELECTION OF CATEGORIES OF STATIONARY SOURCES
Section 111 of the Act directs the Adminstrator to list categories
of stationary sources. The Administrator ". . . shall include a category
of sources in such list if in his judgement it causes, or contributes
significantly to, air pollution which may reasonably be anticipated to
endanger public health or welfare." Proposal and promulgation of
standards of performance are to follow.
Since passage of the Clean Air Amendments of 1970, considerable
attention has been given to the development of a system for assigning
priorities to various source categories. The approach specifies areas
of interest by considering the broad strategy of the Agency for imple-
menting the Clean Air Act. Often, these "areas" are actually pollutants
emitted by stationary sources. Source categories that emit these
pollutants are evaluated and ranked by a process involving such factors
as: (1) the level of emission control (if any) already required by
State regulations, (2) estimated levels of control that might be required
from standards of performance for the source category, (3) projections
of growth and replacement of existing facilities for the source category,
and (4) the estimated incremental amount of air pollution that could be
prevented in a preselected future year by standards of performance for
the source category. Sources for which new source performance standards
were promulgated or under development during 1977, or earlier, were
selected on these criteria.
The Act amendments of August 19/7 establish specific criteria to be
used in determining priorities for all major source categories not yet
listed by EPA. These are: (1) the quantity of air pollutant emissions
that each such category will emit, or will be designed to emit; (2) the
extent to which each such pollutant may reasonably be anticipated to
endanger public health or welfare;-and (3) the mobility and competitive
nature of each such category of sources and the consequent need for
nationally applicable new source standards of performance.
2-5
-------
The Administrator is to promulgate standards for these categories
according to the schedule referred to earlier.
In some cases it may not be feasible immediately to develop a
standard for a source category with a high priority. This might happen
when a program of research is needed to develop control techniques or
because techniques for sampling and measuring emissions may require
refinement. In the developing of standards, differences in the time
required to complete the necessary investigation for different source
categories must also be considered. For example, substantially more
time may be necessary if numerous pollutants must be investigated from a
single source category. Further, even late in the development process
the schedule for completion of a standard may change. For example,
inablility to obtain emission data from well-controlled sources in time
to pursue the development process in a systematic fashion may force a
change in scheduling. Nevertheless, priority ranking is, and will
continue to be, used to establish the order in which projects are
initiated and resources assigned.
After the source category has been chosen, the types of facilities
within the source category to which the standard will apply must be
determined. A source category may have several facilities that cause
air pollution, and emissions from some of these facilities may vary from
insignificant to very expensive to y applying standards to the more
severe pollution sources. For this reason, and because there is no
adequately demonstrated system for controlling emissions from certain
facilities, standards often do not apply to all facilities at a source.
For the same reasons, the standards may not apply to all air pollutants
emitted. Thus, although a source category may be selected to be covered
by a standard of performance, not all pollutants or facilities within
that source category may be covered by the standards.
2-6
-------
2.3 PROCEDURE FOR DEVELOPMENT OF STANDARDS OF PERFORMANCE
Standards of performance must (1) realistically reflect best demon-
strated control practice; (2) adequately consider the cost, the non-air-
quality health and environmental impacts, and the energy requirements of
such control; (3) be applicable to existing sources that are modified or
reconstructed as well as new installations; and (4) meet these conditions
for all variations of operating conditions being considered anywhere in
the country.
The objective of a program for developing standards is to identify
the best technological system of continuous emission reduction that has
been adequately demonstrated. The standard-setting process involves
three principal phases of activity: (1) information gathering,
(2) analysis of the information, and (3) development of the standard of
performance.
During the information-gathering phase, industries are queried
through a telephone survey, letters of inquiry, and plant visits by EPA
representatives. Information is also gathered from many other sources,
and a literature search is conducted. From the knowledge acquired about
the industry, EPA selects certain plants at which emission tests are
conducted to provide reliable data that characterize the pollutant
emissions from well-controlled existing facilities.
In the second phase of a project, the information about the industry
and the pollutants emitted is used 'in an.ilytical studies. Hypothetical
"model plants" are defined to provide a common basis for analysis. The
model plant definitions, national pollutant emission data, and existing
State regulations governing emissions from the source category are then
used in establishing "regulatory alternatives." These regulatory
alternatives are essentially different levels of emission control.
EPA conducts studies to determine the impact of each regulatory
alternative on the economics of the industry and on the national economy,
on the environment, and on energy consumption. From several possibly
applicable alternatives, EPA selects the single most plausible regulatory
alternative as the basis for a standard of performance for the source
category under study.
2-7
-------
In the third phase of a project, the selected regulatory alternative
is translated into a standard of performance, which, in turn, is written
in the form of a Federal regulation. The Federal regulation, when
applied to newly constructed plants, will limit emissions to the levels
indicated in the selected regulatory alternative.
As early as is practical in each standard-setting project, FPA
representatives discuss the possibilities of a standard and the form it
might take with members of the National Air Pollution Control Techniques
Advisory Committee. Industry representatives and other interested
parties also participate in these meetings.
The information acquired in the project is summarized in the Back-
ground Information Document (BID). The BID, the standard, and a preamble
explaining the standard are widely circulated to the industry being
considered for control, environmental groups, other government agencies,
and offices within EPA. Through this extensive review process, the
points of view of expert reviewers are taken into consideration as
changes are made to the documentation.
A "proposal package" is assembled and sent through the offices of
EPA Assistant Administrators for concurrence before the proposed standard
is officially endorsed by the EPA Administrator. After being approved
by the EPA Administrator, the preamble and the proposed regulation are
published in the Federal Register.
As a part of the Federal Register announcement of the proposed
regulation, the public is invited to participate in the standard-setting
process. EPA invites written comments on the proposal and also holds a
public hearing to discuss the proposed standard with interested parties.
All public comments are summarized and incorporated into a second volume
of the BID. All information reviewed and generated in studies in support
of the standard of performance is available to the public in a "docket"
on file in Washington, D. C.
Comments from the public are evaluated, and the standard of performance
may be altered in response to the comments.
2-8
-------
The significant comments and El'A's position on the issues raised
are included in the "preamble" of a "promulgation package," which also
contains the draft of the final regulation. The regulation is then
subjected to another round of review and refinement until it is approved
by the EPA Administrator. After the Administrator signs the regulation,
it is published as a "final rule" in the Federal Register.
2.4 CONSIDERATION OF COSTS
Section 317 of the Act requires an economic impact assessment with
respect to any standard of performance established under Section 111
of the Act. The assessment is required to contain an analysis of
(1) the costs of compliance with the regulation, including the extent to
which the cost of compliance varies depending on the effective date of
the regulation and the development of less expensive or more efficient
methods of compliance, (2) the potential inflationary or recessionary
effects of the regulation, (3) the effects the regulation might have on
small business with respect to competition, (4) the effects of the
regulation on consumer costs, and (5) the effects of the regulation on
energy use. Section 317 also requires that the economic impact assessment
be as extensive as practicable.
The economic impact of a proposed standard upon an industry is
usually addressed both in absolute terms and in terms of the control
costs that would be incurred as a result of compliance with typical,
existing State control regulations. An incremental approach is
necessary because both new and existing plants would be required to
comply with State regulations in the absence of a Federal standard of
performance. This approach requires a detailed analysis of the economic
impact from the cost differential that would exist between a proposed
standard of performance and the typical State standard.
Air pollutant emissions may cause water pollution problems, and
captured potential air pollutants may pose a solid waste disposal problem.
The total environmental impact of an emission source must, therefore, be
analyzed and the costs determined whenever possible.
2-9
-------
A thorough study of the profitability and price-setting mechanisms
of the industry is essential to the analysis so that an accurate estimate
of potential adverse economic impacts can be made for proposed standards.
It is also essential to know the capital requirements for pollution
control systems already placed on plants so that the additional capital
requirements necessitated by these Federal standards can be placed in
proper perspective. Finally, it is necessary to assess the availability
of capital to provide the additional control equipment needed to meet
the standards of performance.
2.5 CONSIDERATION OF ENVIRONMENTAL IMPACTS
Section 102(2)(C) of the National Environmental Policy Act (NEPA)
of 1969 requires Federal agencies to prepare detailed environmental
impact statements on proposals for legislation and other major Federal
actions significantly affecting the quality of the human environment.
The objective of NEPA is to build into the decision-making process of
Federal agencies a careful consideration of all environmental aspects of
proposed actions.
In a number of legal challenges to ..standards of performance for
various industries, the United States Court of Appeals for the District
of Columbia Circuit has held that environmental impact statements need
not be prepared by the Agency for proposed actions under Section 111 of
the Clean Air Act. Essentially, the Court of Appeals has determined
that the best system of emission reduction requires the Administrator to
take into account counter-productive environmental effects of a proposed
standard, as well as economic costs to the industry. On this basis,
therefore, the Court established a narrow exemption from NEPA for EPA
determination under Section 111.
In addition to these judicial determinations, the Energy Supply and
Environmental Coordination Act (ESECA) of 1974 (PL-93-319) specifically
exempted proposed actions under the Clean Air Act from NEPA requirements.
According to section 7(c)(l), "No action taken under the Clean Air Act
shall be deemed a major Federal action significantly affecting the
quality of the human environment within the meaning of the National
Environmental Policy Act of 1969." (15 U.S.C. 793(c)(l))
2-10
-------
Nevertheless, the Agency has concluded that the preparation of
environmental impact statements could have beneficial effects on certain
regulatory actions. Consequently, although not legally required to do
so by section 102(2)(C) of NEPA, EPA has adopted a policy requiring that
environmental impact statements be prepared for various regulatory
actions, including standards of performance developed under section 111
of the Act. This voluntary preparation of environmental impact state-
ments, however, in no way legally subjects the Agency to NEPA requirements.
To implement this policy, a separate section in this document is
devoted solely to an analysis of the potential environmental impacts
associated with the proposed standards. Both adverse and beneficial
impacts in such areas as air and water pollution, increased solid waste
disposal, and increased energy consumption are discussed.
2.6 IMPACT ON EXISTING SOURCES
Section 111 of the Act defines a new source as ". . . any stationary
source, the construction or modification of which is commenced ..."
after the proposed standards are published. An existing source is
redefined as a new source if "modified" or "reconstructed" as defined in
amendments to the general provisions of Subpart A of 40 CFR Part 60,
which were promulgated in the Federal Register on December 16, 1975 (40
FR 58416).
Promulgation of a standard of performance requires States to
establish standards of performance for existing sources in the same
industry under Section 111 (d) of the Act if the standard for new sources
limits emissions of a designated pollutant (i.e., a pollutant for which
air quality criteria have not been issued under Section 108 or which has
not been listed as a hazardous pollutant under Section 112). If a State
does not act, EPA must establish such standards. General provisions
outlining procedures for control of existing sources under Section
lll(d) were promulgated on November 17, 1975, as Subpart R of 40 CFR
Tart. 60 (40 I:R 5.1340).
2-11
-------
2.7 REVISION OF STANDARDS OF PERFORMANCE
Congress was aware that the level of air pollution control achievable
by any industry may improve with technological advances. Accordingly,
Section 111 of the Act provides that the Administrator ". . . shall, at
least every four years, review and, if appropriate, revise . . ." the
standards. Revisions are made to assure that the standards continue to
reflect the best systems that become available in the future. Such
revisions will not be retroactive, but will apply to stationary sources
constructed or modified after the proposal of the revised standards.
2-12
-------
3. DESCRIPTION OF FUGITIVE EMISSION SOURCES
3.1 INTRODUCTION AND GENERAL INDUSTRY INFORMATION
3.1.1 Introduction
The primary purposes of this chapter are to define the synthetic
organic chemical manufacturing industry (SOCMI) and describe the potential
fugitive emission sources that are typically found in this industry. Where
possible, the leak rates of uncontrolled emissions from the various poten-
tial fugitive emission sources are quantified. Industrial practices and
state or local regulations that currently reduce fugitive emissions from
the SOCMI are also briefly discussed in this chapter.
3.1.2 General Information
Organic chemicals are manufactured in a multi-leveled system of
chemical processes that is based on about ten feedstock chemicals which
are principally produced in petroleum refineries. These feedstocks then
proceed through one or more of the process levels and result in literally
thousands of intermediate or finished chemicals (see Figure 3-1).
Generally, each process level contains more chemicals than the preceding
level; the plants manufacturing the products are smaller than the plants
supplying the feedstock; and the volatilities of the products are lower
than the volatilities of the feedstocks. Because of the number and
diverse nature of the organic chemicals included in the multi-leveled
system, the organic chemical industry must be divided into segments for
environmental study and regulation. The synthetic organic chemical
manufacturing industry (SOCMI) is a readily recognizable segment consisting
of some of the higher volume intermediate and finished products. SOCMI
chemicals are the feedstocks for many of the industries producing
synthetic products, such as plastics, fibers, dyes and synthetic rubber.
A list of the SOCMI chemicals is presented in Appendix F.
3-1
-------
RAW MATERIALS
(CRUDE OIL, C3UDE NATURAL GAS. ETC)
REFINERIES
CHEMICAL
FEEDSTOCK
PLANTS
CHEMICAL
FEEDSTOCKS
CHEMICAL
PLANTS
CHEMICAL
PRODUCTS
Figure 3-1. General schematic of process levels that make up
the organic chemical industry.
3-2
-------
Although there are organic chen.ical manufacturing plants in most
industrialized areas of the country, about 60 percent of the SOCMI volume
is produced in Texas and Louisiana. Each plant site may manufacture from
one to several organic chemicals using one or more processes. Although
most processes result in one basic product, some produce a family of
chemicals. Conversely, many chemicals are produced by more than one
process. Yearly, production quantities- at each plant can range from a
few million to several billion kilograms.
3.2 FUGITIVE EMISSION DEFINITION AND POTENTIAL SOURCE DESCRIPTION
3.2.1 Definition
In this study, fugitive emissions in the SOCMI are considered to be
those volatile organic compound (VOC) emissions that result when process
fluid (either liquid or gaseous) leaks from plant equipment. Those VOC
emissions resulting from the transfer, storage, treatment, and/or disposal
of process wastes will be covered by other standards.
3.2.2 Potential Source Characterization and Description
There are many potential sources of fugitive emissions in a typical
synthetic organic chemical plant. The following sources will be con-
sidered in this chapter: pumps, compressors, in-line process valves,
pressure relief devices, open-ended valves, sampling connections, flanges,
agitators, and cooling towers. FugiLive emissions which result from
leaks in these types of equipment anj generally random occurences which
cannot be predicted. Leak occurence is independent of temperature,
pressure, and other process variable', but shows a correlation with vapor
pressure of the substance in the lino. These potential sources are
described below.
3.2.2.1 Pumps. Pumps are used extensively in the SOCMI for the
movement of organic liquids. The centrifugal pump is the most widely
used pump in the SOCMI; however, other types, such as the positive-
diaphragm pumps, are also used in this industry. Chemicals transfered
by pumps can leak at the point of contact between the moving shaft and
stationary casing. Consequently, all pumps except the seal less type
(canned-motor and diaphragm) require a seal at the point where the shaft
penetrates the housing in order to isolate the pump's interior from the
atmosphere.
3-3
-------
Two generic types of seals, packed and mechanical, are currently in
use on pumps in the SOCMI. Packed seals can be used on both reciprocating
and rotary action types of pumps. As Figure 3-2 shows, a packed seal
consists of a cavity ("stuffing box") in the pump casing filled with
special packing material that is compressed with a packing gland to form
a seal around the shaft. Lubrication is required to prevent the buildup
of frictional heat between the seal and shaft. The necessary lubrication
p
is provided by a lubricant that flows between the packing and the shaft.
Deterioration of the packing will result in process liquid leaks.
Fluid
End
Atmosphere
End
1
Possible
/ Lenk
Area
Figure 3-2. Diagram of a simple packed seal.
Mechanical seals are limited in application to pumps with rotating
shafts and can be further categorized as single and dual mechanical
seals. There are many variations to the basic design of mechanical
seals, but all have a lapped seal face between a stationary element and
A
a rotating seal ring. In a single mechanical seal application (Figure 3-3),
the rotating-seal ring and stationary element faces are lapped to a
very high degree of flatness to maintain contact throughout their
entire mutual surface area. The faces are held together by a combination
of pressure supplied by a spring and the pump pressure transmitted
through the liquid which is being pumped. An elastomer seals the rotating
face to the shaft. The stationary face is sealed to the stuffing box
with another elastomer or gasket.
3-4
-------
PUMP
STUFFING
BOX
SHAFT
.— STATIONARY
ELEMENT
=OSSIBLE
LEAK AREA
Figure 3-3. Diagram of a basic single mechanical seal/
In a dual mechanical seal application, two seals are usually arranged
back-to-back or in tandem. In the back-to-back arrangement (Figure 3-4),
the two seals provide a closed cavity between them. A seal liquid, such
as water or seal oil, is circulated through the cavity. Because the
seal liquid surrounds the two seals, it can be used to control the
temperature in the stuffing box. In order for the seal to function, the
seal liquid must be at a pressure greater than the operating pressure of
the stuffing box. As a result, any leakage would be across the seal
faces. Liquid leaking across the inboard face would enter the stuffing
box and mix with the process liquid. Seal liquid going across the
outboard face would exit to the atmosphere. Therefore, the seal liquid
must be compatible with the process liquid as well as with the environment.1
In a tandem dual mechanical seal arrangement (Figure 3-5), the
seals face the same direction. The secondary seal provides a backup for
the primary seal. The cavity between the two seals is filled with a
buffer liquid which may be used for temperature control in the stuffing
box. However, the barrier liquid may be at a pressure lower than that
in the stuffing box. Therefore, any leakage would be from the stuffing
3-5
-------
POSSIBLE LEAK
INTO SEALING
FLUID
FLUID END
GLAND
PLATE
PRIMARY —i
SEAL
V
SECONDARY
SEAL
Figure 3-4. Diagram of a dual mechanical seal
(back-to-hack arrangement).
PRIMARY
SEAL
BUFFER LI QUID
OUT , IN
(TOP) (BOTTOM)
V
SECONDARY
SEAL
GLAND
PLATE
70-1787.1
Figure 3-5. Diagram of a dual mechanical seal
(tanden arrangement).
3-6
-------
box into the barrier liquid. Since this liquid is routed to a closed
reservoir, process liquid that has leaked into the seal cavity will also
be transferred to the reservoir. At the reservoir, the process liquid
could vaporize and be emitted to the atmosphere. To ensure that VOC's
do not leak from the reservoir, the reservoir can be vented to a control
9
device.
Another arrangement of dual seals which represents a relatively
new development is the face-to-face arrangement. In this configuration
two rotating faces are mated with a common stationary. Barrier fluid
may be provided at higher or lower pressures than the stuffing box. As
in the tandem arrangement, if the barrier fluid is at a lower pressure
than the stuffing box, the barrier fluid reservoir would require venting
to a control device.
Another type of pump that has been used in the chemical industry is
the seal less pump. Canned-motor and diaphragm pumps are seal less pumps.
In the canned-motor pumps the cavity housing the motor rotor and the
pump casing are interconnected. As a result, the motor bearings run in
the process liquid and all seals are eliminated. Because the process
liquid is the bearing lubricant, abrasive solids cannot be tolerated.
Canned-motor pumps are being widely used for handling organic solvents,
organic heat transfer liquids, light oils, as well as many toxic or
hazardous liquids, or where leakage is an economic problem.
Diaphragm pumps (see Figure 3-6) perform similarly,, to piston and
plunger pumps. However, the driving member is a flexible diaphragm
L.<:;CHARGE
CHECK VALVE \
INLET
'CHECK VALVE
/— DIAPHRAGM
PISTON
1 °
Kigure 3-f>. Diaphragm Pump1*1
3-7
-------
fabricated of metal, rubber, or plastic. The primary advantage of this
arrangement is the elimination of all packing and seals exposed to the
process liquid. This is an important asset when hazardous or toxic
liquids are handled.
3.2.2.2 Compressors. Gas compressors used in the SOCMI are similar
to pumps in that they can be driven by rotary or reciprocating shafts.
They are also similar to pumps in their need for shaft seals to isolate
the process gas from the atmosphere. As with pumps, these seals are likely
to be the source of fugitive emissions from compressors.
Shaft seals for compressors may be chosen from several different
types: labyrinth, restrictive carbon rings, mechanical contact, and
liquid film. All of these seal types are leak restriction devices; none
of them completely eliminatesleakage. Many compressors may be equipped
with ports in the seal area to evacuate gases collecting there.
The labyrinth type of compressor seal is composed of a series of
close tolerance, interlocking "teeth" which restrict the flow of gas along
the shaft. A straight pass labyrinth compressor seal is shown in Figure 3-7.
Many variations in "tooth" design and materials of construction are
available. Although labyrinth type seals have the largest leak potential
of the different types, properly applied variations in "tooth" configuration
and shape can reduce leakage by up to 40 percent over a straight pass type
labyrinth.13
Restrictive0carbon ring seals consist of multiple stationary carbon
rings with close shaft clearances. This type of seal may be operated dry
with a sealing fluid or with a buffer gas. Restrictive ring seals can
14
achieve lower leak rates than the labyrinth type. A restrictive ring
seal is shown in Figure 3-8.
Mechanical contact seals (shown in Figure 3-9) are similar to the
mechanical seals described for pumps. In this type of seal clearance
between the rotating and stationary elements is reduced to essentially zero.
011 or another suitable lubricant is supplied to the seal faces. Mechanical
contact seals can achieve the lowest leak rates of the types described
15
here, but they are not suitable for all processing conditions.
Centrifugal compressors also can be equipped with liquid film seals.
A diagram of a liquid film seal is shown in Figure 3-10. The seal is
formed by a film of oil between the rotating shaft and stationary gland.
3-8
-------
PORT MAY BE ADDED
FOR SCAVENGING OH
INERT-GAS SEALJNG —-
WvvvvvvvW'"Vi
INTERNAL I
GAS PRESSURED-
ATMOSPHERE
Figure 3-7. Labyrinth compressor seal.*
PORT MAY BE
ADDED FOR
SEALJNG
SCAVENGING
PORT MAY BE
ADDED ?OR
VACUUM
APPLICATION
Figure 3-8. Restrictive ring compressor seal.*
*American Petroleum Institute. Centrifugal Compressors for Refinery Service,
API Standard 617, 4th ed., pp. 8-9. Reprinted by Courtesy of the American
Petroleum Institute.
3-9
-------
INTERNAL
GAS PRESSURE
CLEAN OIL IN
PRESSURE
BREAKDOWN
SLEEVE
STATIONARY SEAT •
CARBON RING
CONTAMINATED
OIL OUT
Figure 3-9. Mechanical contact compressor seal.*
- CLEAN OIL IN
ATMOSPHERE
CONTAMINATED
OIL OUT
OIL OUT
Figure 3-10. Liquid film compressor seal.*
*Amefican Petroleum Institute. Centrifugal Compressors for Refinery Service,
API Standard 617, 4th ed., pp. 8-9. Reprinted by Courtesy of the American
Petroleum Institute.
3-10
-------
When the circulating oil is returned tc the oil reservoir, process gas can
be released to the atmosphere.^ To eliminate release of VOC emissions
from the seal oil system, the reservoir can be vented to a control device.
3.2.2.3 Process Valves. One of the most common pieces of equipment
in organic chemical plants is the valve. The types of valves commonly
used are control, globe, gate, plug, ball, relief, and check valves. All
except the relief valve (to be discussed further below) and check valve
are activated by a valve stem, which may have either a rotational or
linear motion, depending on the specific design. This stem requires a
seal to isolate the process fluid inside the valve from the atmosphere as
illustrated by the diagram of a gate valve in Figure 3-11. The possibility
of a leak through this seal makes it a potential source of fugitive
emissions. Since a check valve has no stem or subsequent packing gland,
it is not considered to be a potenlial source of fugitive emissions.
Sealing of the stem to prevent leakage can be achieved by packing
inside a packing gland or 0-ring seals. Valves that require the stem to
move in and out with or without rotation must utilize a packing gland.
Conventional packing glands are suited for a wide variety of packing
material; the most common are various types of braided asbestos that
contain lubricants. Other packing materials include graphite, graphite-
impregnated fibers, and tetrafluorethylene; the packing material used
depends on the valve application and configuration. These conventional
packing glands can be used over a wide range of operating temperatures.
At high pressures these glands must be quite tight to attain a good seal.
18
Figure 3-11.. Diagram of a gate valve:
19
3-11
-------
Elastomeric 0-rings are also used for sealing process valves. These
0-rings provide good sealing but are not suitable where there is sliding
motion through the packing gland. Those seals are rarely used in high
pressure service and operating temperatures are limited by the seal
material.20
BelTows seals are more effective for preventing process fluid leaks
than the conventional packing gland or any other gland-seal arrangement.^
This type of seal incorporates a formed metal bellows that makes a barrier
.between the disc and body bonnet joint. An example of this seal is
presented in Figure 3-12. The bellows is the weak point of the system
and service life can be quite variable. Consequently, this type of seal
is normally backed up with a conventional packing gland and is often fitted
with a leak detector in case of failure. ^
BELLOWS
BODYBONNET
Figure 3-12. ExampU of bellows seals.23
A diaphragm may be used to isolate the working parts of the valve and
the environment from the process liquid. Two types of valves which utilize
diaphragms are illustrated in Figures 3-11(a) and (b). As Fiqure 3-11(b)
shows, the diaphragm may also be used to control the flow of the process
fluid. In this design, a compressor component pushes the diaphragm toward
the valve bottom, throttling the flow. The diaphragm and compressor are
connected in a manner so that it is impossible for them to be separated
under normal working conditions. When the diaphragm reaches the valve
3-12
-------
bottom, it-seats firmly against the bottom, forming a leak-proof seal.
This configuration is recommended for fluids containinq solid particles
and for medium-pressure service. Depending on the diaphragm material,
this type of valve can be used at temperatures up to 205°C and in severe
acid solutions. If failure of the seal occurs, a valve employinq a dia-
phragm seal can become a source of fugitive emissions.
DIAPHRAGM
DISK —
STEM
DIAPHRAGM
!0 1/71 I
Figure 3-13. Diagrams of valves with diaphragm seals.
25
3.2.2.4 Pressure Relief Devices. Engineering codes require that
pressure-relieving devices or systems be used in applications where the
process pressure may exceed the maximum allowable working pressure of the
vessel. The most common type of pressure-relieving device used in the
SOCMI is the pressure relief valve (Figure 3-14). Typically, relief valves
are spring-loaded and designed to open when the process pressure exceeds a
set pressure, allowing the release of vapors or liquids until the system
pressure is reduced to its normal operating level. When the normal
3-13
-------
pressure is re-attained, the valve reseats, and a seal is again formed.
The seal is a disk on a seat, and the possibility of a leak through this
seal makes the pressure relief valve a potential source of VOC fugitive
emissions. Two potential causes of leakage from relief valves are:
"simmering or popping", a condition due to the system pressure being
close to the set pressure of the valve, and improper reseating of the
97
valve after a relieving operation. '
Rupture disks are also common in the SOCMI. These disks are made of
a material that ruptures when a set pressure is exceeded, thus allowing
the system to depressurize. The advantage of a rupture disk is that the
disk seals tightly and does not allow any VOC's to escape from the system
under normal operation. However, when the disk does rupture, the system
depressurizes until atmospheric conditions are obtained; this could result
in an excessive loss of product or correspondingly an excessive release
of fugitive emissions.
Possible
Leak
Area
Process Side
Figure 3-1 4. Diagram of a spring-loaded relief valve.
3-14
-------
3.2.2.5 Cooling Towers. Cooling towers are found in most SOCMI
plants. The purpose of these towers is to cool the plant's process cooling
waters which have been heated while removing heat from various process
equipment (reactors, condensers, heat exchangers). This cooling process
is achieved by evaporation when the process cooling water and air are
contacted. Under normal operating conditions, a cooling tower would not
be considered a fugitive emission source. However, if a leak occurs in
the process equipment and if this equipment is operating at a pressure
greater than that of the cooling water, organic chemicals can leak into the
water. When the process water is recirculated to the cooling tower, these
chemicals can be released to the atmosphere.^
3.2.2.6 Agitators. Agitators are commonly used in the SOCMI to
stir or blend chemicals. Like pumps and compressors, agitators may
leak organic chemicals at the point where the shaft penetrates the casing.
Consequently, seals are required to minimize fugitive emissions from
agitators. Four seal arrangements are commonly used with agitators; they
include: compression packing (packed seal), mechanical seals, hydraulic
29
seals, and lip seals. Packed seals for agitators are very similar in
design and application to the packed seals for pumps (Section 3.2.2.1).
Although mechanical seals are more costly than the other three seal
arrangements, they offer a greatly reduced leakage rate to offset their
higher cost. The maintenance frequency of mechanical seals is, also, one-
half to one-fourth that of packed seals.3C In fact, at pressures greater
than 1140 kPa (150 psig), the leakage rate and maintenance frequency are
so superior that the use of packed seals on agitators is rare.3i £s with
packed seals, the mechanical seals for agitators are similar to the design
and application of mechanical seals for pumps (Section 3.2.2.1).
The hydraulic seal (Figure 3-l'.>) is the simplest and least used
agitator shaft-seal. In this type of seal, an annular cup attached to the
process vessel contains a liquid that is in contact with an inverted cup
3-15
-------
attached to the rotating agitator shaft. The primary advantage of this
seal is that it is a non-contact seal. However, this seal is limited to
low temperatures and pressures and can only handle very small pressure
fluctuations. Organic chemicals may contaminate the seal liquid and then
be released into the atmosphere as fugitive emissions.3^
INVERTED CUP
ANNULARCUP
70-1772-1
oo
Figure 3-15. Diagram of hydraulic seal for agitators.
A lip seal (Figure 3-16) can be used on a top-entering agitator as a
dust or vapor seal. The sealing element is a spring-loaded elastomer.
Lip seals are relatively inexpensive and easy to install. Once the.seal
has been installed the agitator shaft rotates in continuous contact with
the lip seal. Pressure limits of the seal are 2 to 3 psi because it
operates without lubrication. Operating temperatures are limited by the
characteristics of the elastomer. Fugitive VOC emissions could be
released through this seal when this seal wears excessively or the
operating pressure surpasses the pressure limits of the
Figure 3-16. Diagram of agitator lip seal.
3-16
35
-------
3.2.2.7 Open-Ended Valves or Lines. Some valves are installed in a
system so that they function with the downstream line open to the atmos-
phere. Examples are purge valves, drain valves, and vent valves. A
faulty valve seat, or incompletely closed valve would result in leakage
through the valve and fugitive VOC emissions to the atmosphere.
3.2.2.8 Sampling Connections. The operation of a process unit is
checked periodically by routine analyses of feedstocks and products. To
obtain representative samples for these analyses, sampling lines must
first be purged prior to sampling. The purged liquid or vapor is
sometimes drained onto the ground or into a sewer drain, where it can
evaporate and release VOC emissions to the atmosphere.
3.2.2.9 Flanges. Flanges are bolted, gasket-sealed junctions used
wherever pipe or other equipment such as vessels, pumps, valves, and heat
exchangers may require isolation or removal. Normally, flanges are
employed for pipe diameters of 50 mm or greater and are classified by
pressure and face type.
^Flanges may become fugitive emission sources when leakage occurs due
to improperly chosen gaskets or a poorly assembled flange. The primary
cause of flange leakage is due to thermal stress that piping or flanges in
some services undergo; this results in the deformation of the seal between
the flange faces. 36
3.3 BASELINE CONTROL
There are presently no federal regulations that specifically reduce
emissions from synthetic organic chemical manufacturing plants. However,
some fugitive emission reduction is achieved by operating practices
currently followed by industry and applicable state or local regulations.
Because these practices and regulations only "incidentally" control
fugitive emissions, they are considered, in this study, to be the baseline
control level. The procedures, specific control techniques, and regula-
tions that make up the baseline control level are discussed below.
3-17
-------
Fugitive emissions occurring under the baseline control level are
subsequently considered in this report to be uncontrolled emissions. Data
characterizing the uncontrolled levels of fugitive emissions in the SOCMI
are presently unavailable. However, data of this type have been obtained
for the refining industry. These data are presented in Table 3-1.
Because the operation of the various process equipment in the SOCMI is not
expected to differ greatly from the operation of the same equipment in
the refining industry, it is felt that the refinery fugitive emission data
can be used to approximate the levels of fugitive emissions in SOCMI. Test
data in Appendix C.I indicate that this engineering judgement is reasonable.
These data show that leak rates and leak frequencies within SOCMI and
petroleum refineries are similar.
3.3.1 Industrial Practices
The organic chemical industry has been primarily interested in leaks
that are large enough to be physically evident (leaks that can be seen,
heard, or smelled); such leaks are termed "easily detectable leaks" and
are normally repaired to minimize the loss of product. Fugitive emissions,
as they are considered in this report, have considerably smaller emission
rates than "easily detectable leaks." In the past, SOCMI generally has
not monitored equipment for fugitive emissions nor repaired equipment
on the basis of reducing the level of fugitive emissions. Processes
which have emitted toxic or hazardous compounds have been exceptions to
this rule.
While SOCMI has been concerned primarily with easily detectable
leaks, certain equipment and procedures used in many organic chemical
plants may help to reduce fugitive VOC emissions. For instance, some
plants cap-off or use double block valves on the end of process lines.
Either of these procedures will reduce fugitive emissions. In some plants
relief valves are checked to see if the valve has reseated properly after
27
relieving. As previously mentioned, an improperly seated relief valve
may allow fugitive VOC emissions to orcur. Rupture discs, which are
commonly used in the SOCMI, also prevent fugitive VOC emissions. Some
organic chemical plants employ closed-loop sampling which help to reduce
fugitive emissions.
•3-18
-------
TABLE 3-1. UNCONTROLLED HUilllVE ll'.ISSI.'N FACTORS IN I III-: SYN1IIHK,
ORGANIC CHEMICAL HANUI A( HIKING INDUViKY (SOCMI)
Uncontrolled emission
Fugitive emission source factor,3 kg/hr
Pumps
Light liquids
With packed seals 0.12
With single mechanical seals 0.12
With double mechanical seals 0.12C
With no seals 0.0
Heavy Liquids
With packed seals 0.020
With single mechanical seals 0.020
With double mechanical seals 0.020
With no seals 0.0
Valves (in-line)
Gas . 0.021
Light liquid° 0.010
Heavy liquid0 0.0003
Safety/relief valves
Gas . 0.16
Light liquid" 0.006
Heavy liquid 0.009
Open-ended valves
Gas . 0.025
Light liquid? 0.014
Heavy liquid0 0.003
Flanges • 0.0003
Sampling connections 0.015
Compressors 0.44
Cooling towers 13.6-11076
Agitators
a
'These uncontrolled emission levels are based upon the refinery data presented
in reference 38.
bLight liquid is defined as a fluid with vapor pressure greater than 0.3 kPa
at 20°C. This vapor pressure represents the split between kerosene and naphtha
and is based on data presented in reference 39. The average vapor pressure of
liquids falling between these two components is approximately 0.04 psi at 68°F.
cAssumes the inner seal leaks at the same rate as single seal and that the VOC
is emitted from the seal oil degassing vent.
dHeavy liquid is defined as a fluid with vapor pressure less than 0.3 kPa at
20°C. This vapor pressure represents the split between kerosene and naphtha
and is based on data presented in reference 40. The average vapor pressure of
liquids falling between these two components is approximately 0.04 psi at 68°F.
eThese levels are based on cooling tower circulation rates that range from
0.05-3.66 m3/sec (714-58,000 GPM). Ref. 41.
NA = no data available.
3-19
-------
The flaring of vapors vented from various vessels or equipment is
another technique which is used by some plants (particularly those producing
toxic or hazardous chemicals) to reduce fugitive emissions.
3.3.2 Existing Regulations
There are, presently, two types of regulations that impact fugitive
VOC emissions from organic chemical plants. The first type is to regulate
industrial operating practices on the basis of worker health and safety.
Because some aspects of these regulations deal with worker exposure to
process emissions, they may have some impact on fugitive VOC emissions.
The second type of regulations is regulations that were specifically de-
veloped to limit fugitive emissions.
3.3.2.1 Health and Safety Regulations. Several regulations have
been established under the direction of the Occupational Safety and Health
Administration and National Institute for Occupational Safety and Health
to limit worker exposure to chemical substances. Protecting the workers
may be accomplished by either limiting the level of emissions or by
providing workers with protection from the emissions. In this way,
regulations may result in a reduction in the levels of fugitive VOC
emissions.
In the vinyl chloride monomer and benzene industries, safety
and health regulations are designed to limit the ambient VOC levels
to which workers may be exposed. Since1 t.hesi; standards do not stipulate
how the allowable ambient levels should be achieved, workers can be
protected from high ambient VOC levels by: 1) a reduction in the fugitive
VOC emissions or 2) the use of special equipment (such as personal
respirators) to isolate the worker from the emissions. This example
illustrates that the present health and safety regulations do not •;
mandate a reduction in fugitive VOC emissions, and any reduction in
.fugitive emissions' resulting from'-these regulations can be considered
to be "incidental". By contrast, fugitive emission regulations do
require the fugitive emissions to be reduced. .
3-20
-------
3.3.2.2 Fugitive Emissions Regulations. Currently, there are no
federal fugitive emission regulations for the SOCMI. However, California
has established such regulations, and organic chemical plants in this
state must comply with the approoriate regulations.
California presently requires open-ended process lines to be capped-off
in order to minimize fugitive VOC emissions. This state also requires
relief valves to be vented to a flare system, monitored and maintained, or
a rupture disk to be used. In addition to these regulations, the South
Coast Air Quality Management District requires organic chemical plants
to vent fugitive emissions from compressor seals to a fired-heater or
flare system. The South Coast and Bay Area AQMD also require periodic
inspection of valves in the chemical and refining industries.
3-21
-------
3.4 REFERENCES
1. Erikson, D. G. and V. Kalcevic. (Hydroscience, Inc.) Emissions
Control Options for the Synthetic Organic Chemicals Manufacturing
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577. February
1979. p. II-2.
2. Reference 1.
3. Reference 1, p. 11-3.
4. Ramsden, J. H. How to Choose and Install Mechanical Seals. Chemical
Engineering. 85(22):97-102. October 9, 1978.
5. Reference 1, p. 11-3.
6. Reference 4, p. 99.
7. Reference 4, p. 100.
8. Reference 4, p. 101.
9. Reference 4, p. 99.
10. Perry, R. H. and C. H. Chilton. Chemical Engineers' Handbook,
Fifth Edition. New York, McGraw-Hill Book Company, 1973. p. 6-8.
11. Reference 10, p. 6-13.
12. Birk, J. R. and J. H. Peacock. Pump Requirements for the Chemical
Process Industries. Chemical Engineering. {Jl_(4):120. February 18,
1974.
13. Nelson, W. E. Compressor Seal Fundamentals. Hydrocarbon Processing.
56(12):91-95. December 1977.
14. Reference 13.
15. Reference 13.
16. Reference 1, p. 11-7.
17. Lyons, J. L. and C. L. Askland. Lyons' Encyclopedia of Valves.
New York, Van Nostrand Reinhold Company, 1975. 290 p.
18. Templeton, H. C. Valve Installation, Operation and Maintenance.
Chemical Engineering. _7^(23):141-149. October 11, 1971.
19. Reference 1, p. II-5.
3-22
-------
20. Reference 18, p. 147-148.
21. Reference 18, p. 148.
22. Reference 18, p. 148.
23. Reference 18, p. 148.
24. Pikulik, A. Manually Operated Valves. Chemical Engineering.
£5(7):121. April 3, 1978,
25. Reference 24, p. 121.
26. Steigerwald, B. J. Emissions of Hydrocarbons to the Atmosphere
from Seals on Pumps and Compressors. (Prepared for the Joint
District, Federal and State Project for the Evaluation of Refinery
Emissions.) Report No. 6. April 1958. 37 p.
27. Reference 1, p. II-7.
28. Cooling Tower Fundamentals and Application Principles. Kansas
City, The Marley Company, 1969. p. 4.
29. Ramsey, W. D. and G. C. Zoller. How the Design of Shafts, Seals
and Impellers Affects Agitator Performance. Chemical Engineering.
83(18):101-108. August 30, 1976.
30. Reference 29, p. 105.
31. Reference 29, p. 105.
32. Reference 29, p. 105.
33. Reference 29, p. 106.
34. Reference 29, p. 106.
35. Reference 29, p. 106.
36. McFarland, I. Preventing Flange Fires. Chemical Engineering
Progress. j>5_(8) :59-61. August 1969.
37. Letter and attachments from Johnson, J. M., Exxon Company, To
Walsh, R. T., EPA:CPB. July 28, 1977. 14 p. Review of "Control
of Hydrocarbon from Miscellaneous Refinery Sources" report.
3-23
-------
38. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979. p. 22.
39. Reference 38.
40. Reference 38.
41. Radian Corporation. Assessment of Atmospheric Emissions from
Petroleum Refining, Appendix B: Detailed Results. (Prepared for
U. S. Environmental Protection Agency.) Research Triangle Park,
N. C. Publication No. EPA-600/2-80-075C. April 1980.
3-24
-------
4. EMISSION CONTROL TECHNIQUES
Sources of fugitive VOC emissions from SOCMI plants were identified
in Chapter 3 of this document. The potential emission control techniques
that can be applied to SOCMI fugitive emission sources are discussed in
this chapter. The applicability and estimated control effectiveness of
each technique are also presented. The quantitative control effective-
ness for many of the control techniques is not known. Qualitative
discussions of effectiveness and references to technology transfer from
similar industries are presented wherever applicable.
4.1 LEAK DETECTION AND REPAIR METHODS
Leak detection and repair methods can be applied in order to reduce
fugitive emissions from any source. Leak detection methods are used to
identify equipment components that are emitting significant amounts of
VOC. Emissions from leaking sources may be reduced by three general
methods: repair, modification, or replacement of the source.
4.1.1 Leak Detection Methods
Leak detection methods include individual component surveys, area
(walk-through) surveys, and fixed point monitors. The first method
(individual component surveys) is also a part of the other methods.
4.1.1.1 Individual Component Survey. Each fugitive emission source
(pump, valve, compressor, etc.) is checked for VOC leakage in an individ-
ual component survey. The source may be checked for leakage by visual,
audible, olfactory, soap bubble, or instrument techniques. Visual methods
are good for locating liquid leaks, especially pump seal failures.
Observation of a visible leak does not necessarily indicate VOC emissions,
since the leak may be composed of non-VOC compounds. High pressure leaks
4-1
-------
may be detected by hearing the escaping vapors, and leaks of odorous
materials may be detected by smelling the odor. Current industry
practices include leak detection by visual, audible, and olfactory
methods. However, in many instances, even very large VOC leaks are not
detected by these methods.-
Spraying soap on equipment components is another individual survey
method. If the soap solution forns bubbles or is blown away, a leak from
the component is indicated. A disadvantage of this method is that it does
not distinguish leaks of non-VOC compounds from VOC leaks. Consequently,
air or steam leaks would produce the same observed effect as VOC leaks.
This method is only semiquantitative since it requires that the observer
subjectively determine the rate of leakage based on behavior of the soap
bubbles. This method is limited to "cool" sources, since temperatures
above 100°C would cause the water in the soap solution to boil away. This
method is also not suited for moving shafts on pumps or compressors, since
the motion of the shaft may interfere with the motion of the bubbles caused
by a leak.
Portable hydrocarbon detection instruments are the best method for
identifying leaks of VOC from equipment components. The instrument is
used to sample and analyze the air in close proximity to the potential
leak surface by traversing the sampling probe tip over the entire area
where leaks may occur. This sampling traverse is called "monitoring" in
subsequent descriptions. The hydrocarbon concentration of the sampled air
is displayed on the instrument meter. The performance criteria for moni-
toring instruments and a description of instrument survey methods are
included in Appendix D. The hydrocarbon concentration observed during
monitoring of a component is proportional to the VOC emission rate from
the component. Data from petroleum refineries have been used to develop
relationships between monitoring concentration and mass emission rates.
The hydrocarbon concentration which indicates that a component needs mainte-
nance must be chosen. Components which have indicated concentrations
higher than this "action level" are marked for repair. Data from
petroleum refineries indicate thai large variations in mass emission rate
4-2
-------
may occur over short time periods for an individual equipment component.
More frequent monitoring intervals tend to reduce the chance of missing
"large leaks" because of their variable leak rates.
4.1.1.2 Area Survey. An area survey (also known as a walk-through
survey) requires the use of a portable hydrocarbon detector and a strip
chart recorder. The procedure involves carrying the instrument within one
meter of the upwind and downwind sides of process equipment and associated
fugitive emission sources. An increase in observed concentration indi-
cates leaking fugitive emission sources. The instrument is then used for
an individual component survey in the suspected leak area. The efficiency
of this method for locating leaks is not well established. It has been
estimated that the walk-through survey combined with selected individual
surveys will detect about 50 percent of the number of leaks identified in a
complete individual survey. The time and labor requirements for the
walk-through are much lower. This method will not detect leaks from
sources such as elevated valves or relief valves. Leaks from adjacent •'
units and adverse meteorological conditions can affect the results of the
walk-through survey. Consequently, the walk-through survey is best for
locating only large leaks with a small resource expenditure.
4.1.1.3 Fixed Point Monitors. This method consists of placing
several automatic hydrocarbon sampling and analysis instruments at
various locations in the process unit. The instruments may sample the
ambient air intermittently or continuously. Elevated hydrocarbon concen-
trations indicate a leaking component. As in the walk-through method, an
individual component survey is required to identify the specific leaking
component in the area. For this method, the portable hydrocarbon detec-
tor is also required. Leaks from adjacent units and meteorological
conditions may affect the results obtained. The efficiency of this
method is not well established, but it has been estimated that 33 percent
of the number of leaks identified by a complete individual component
survey could be located by fixed-point monitors.^ Fixed-point monitors
are more expensive, multiple units may be required, and the portable
instrument is also required to locate the specific leaking component.
4-3
-------
Calibration and maintenance costs may be higher. Fixed-point monitors have
been used to detect emissions of hazardous or toxic substances (such as vinyl
chloride) as well as potentially explosive conditions. Fixed-point monitors
have an advantage in these cases, since a particular compound can be selected
as the sampling criterion.
4.1.2 Repair Methods
The following descriptions of repair methods include only those
features of each fugitive emission source (pump, valve, etc.) which need
to be considered in assessing the applicability and effectiveness of each
method. They are not intended to be complete repair procedures. The
effectiveness of repairs in reducing fugitive emissions has not been
well documented; however, data for valve repairs have been collected in
various petroleum refineries. In many cases, perfect repair will not be
achieved, but whenever repairs are performed, the portable hydrocarbon
detector should be used to identify the lowest achievable emission rate.
4.1.2.1 Pumps. Many pumps have spares which can be operated while
the leaking pump is being repaired. Leaks from packed seals may be reduced
by tightening the packing gland. At some point, the packing may deteriorate
to the point where further tightening would have no effect or possibly even
increase fugitive emissions from the seal. The packing can be replaced with
the pump out of service. When mechanical seals are utilized, the purnp must
be dismantled so the leakino seal can be repaired or replaced. Dismantling
pumps, 1f the seal leak is small, may result in spillane of some process
fluid and evaporative emissions of VOC. These temporary emissions may be
greater than the continued leak from the seal.
4.1.2.2 Compressors. Leaks from packed seals may be reduced by the
same repair procedure that was described for pumps. Other types of seals
require that the compressor be out of service for repair. Since most compressors
do not have spares, repair or replacement of the seal would require a shut-
down of the process. If the leak is small, temporary emissions resulting
from a shutdown may be greater than the emissions from the leaking seal.
4-4
-------
~~ """4.1.2.3 Rel ief Valves. In general, relief valves which leak must
be removed in order to repair the leak. In some cases of improper reseat-
ing, manual release of the valve may improve the seat seal. In order to
remove the relief valve without shutting down the process, a block valve may
be required upstream of the relief valve. A spare relief valve should be
attached while the faulty valve is repaired and tested. After a relief
valve has been repaired and replaced, there is no guarantee that the
next over-pressure relief will not result in another leak.
4.1.2.4 Valves. Most valves have a packing gland which can be
tightened while in service. Although this procedure should decrease the
emissions from the valve, in some cases it may actually increase the
emission rate if the packing is old and brittle or has been overtightened.
Plug type valves can be lubricated with grease to reduce emissions around
the plug. Some types of valves have no means of in-service repair and
must be isolated from the process and removed for repair or replacement.
Other valves, such as control valves, may be excluded from in-service
repair by operating or safety procedures. In many cases, valves cannot
be isolated from the process for removal. Most control valves have a
manual bypass loop which allows them to be isolated and removed. Most
block valves cannot be isolated easily although temporary changes in
process operation may allow isolation in some cases. If a process unit
must be shut down in order to isolate a leaking valve, the emissions
resulting from the shutdown will probably be greater than the emissions
from the valve if allowed to leak until the next process change which
permits isolation for repair.
Depending on site specific factors, it may be possible to repair process
valves by injection of a sealing fluid into the source. This type of repair
may affect the operability of the valve so that replacement of the source
might be necessary within a short time after its repair. Injection of
sealing fluid has been successfully used to repair leaks from valves in
q
petroleum refineries in California.
4-5
-------
4.1.2.5 Flanges. In some cases, leaks from flanges can be reduced
by replacing the flange gaskets. Most flanges cannot be isolated to
permit replacement of the gasket. Data from petroleum refineries show
that flanges emit very small amounts of VOC.4
4.1.3 Control Effectiveness of Leak Detection and Repair Methods
The instrument survey of individual components is the only type of leak
detection method for which control effectiveness has been quantified.
The following estimations of control effectiveness do not pertain to the
soap bubble leak detection method, area surveys, or fixed-point monitoring
methods.
There are several factors which determine the control effectiveness of
individual component surveys; these include
Action level or leak definition,
Inspection interval or monitoring frequency,
Achievable emission reduction of maintenance, and
Interval between detection and repair of the leak.
Some of these factors can be estimated by using data collected from
petroleum refineries.5
4.1.3.1 Action Level. The action level is the minimum hydrocarbon
concentration observed during monitoring which defines a leaking component
which requires repair. The choice of the action level for defining a
leak is influenced by a number of important considerations. First, the
percent of total mass emissions which can potentially be controlled by
the monitoring and repair program can be affected by varying the leak
definition, or action level. Table 4-1 gives the percent of total mass
emissions affected by various action levels for a number of equipment
types. The data in this table, indicate that, in general, a low action
level results in larger potential emission reductions. However, the
choice of an appropriate leak definition is most importantly limited by
the ability to repair leak ing component1;. Test data indicate that about
90 percent of valve leaks with initial screening values equal to or greater
than 10,000 ppmv can be successfully repaired (see Appendix C). Similar
data indicate that attempted repair of valve leaks with initial screening
4-6
-------
TABLE 4-1. FRACTION OF TOTAL MASS EMISSIONS FROM VARIOUS SOURCE TYPES
THAT WOULD BE AFFECTED BY DIFFERENT ACTION LEVELS
Action level (ppmv)
Source type
Pump seals
Light liquid service
Heavy liquid service
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Compressor seals
Flanges
Fraction of mass emissions (as %J^
100,000
56
0
85
49
0
20
28
0
50,000
68
0
92
62
0
33
48
0
10,000
87
21
98
84
0
69
84
0
1,000
97
66
99
96
23
92
98
48
These data show the fraction of the total emissions from a given source
type that is attributable to sources with leaks above the various action
levels.6
Level of emission at which repair of the source is required.
4-7
-------
values of less than 10,000 ppmv can increase instead of decrease emissions
from these valves. From these data it is concluded that repairing leaks
with screening values in the 1,000-10,000 ppmv range may not result in a
net reduction in mass emissions.7 The nature of repair techniques for
pipeline valves, for instance, are such that to repair leaks below
a certain level by tightening valve packing may actually result in an
increase in emissions. In practice, valve packing material becomes hard
and brittle after extended use. As the packing loses its resiliency, the
valve packing gland must be tightened to prevent loss of product to the
atmosphere. Excessive tightening, however, may cause cracks in the packing,
thereby increasing the leak rate. Unbalanced tightening of the packing gland
may also cause the packing material to be positioned improperly in the valve
and allow leakage. Valves which are not often used can build up a "static"
seal of paint or hardened lubricant which could be broken by tightening
the packing gland. Therefore, it may be important not to cause small
leaks to become large leaks by requiring tightening of valves to meet a
very low leak repair action level.
4.1.3.2 Inspection Interval. A monitoring plan may include annual,
quarterly, monthly, or even weekly inspections. The length of time
between inspections should depend on the expected occurrence and
recurrence of leaks after a piece of equipment has been checked or
repaired. This interval can be related to the type of equipment and
service conditions, and different intervals can be specified for different
pieces of equipment after appropriate equipment histories have been
developed. In the refinery VOC leak Control Techniques Guideline (CTG)
document, the recommended monitoring intervals are:- annual--pump seals,
pipeline valves in liquid service, and process drains; quarterly--
compressor seals, pipeline valves in gas service, and pressure relief
valves in gas service; weekly—visual inspection of pump seals; and no
individual monitoring—pipeline flanges and other connections, and
pressure relief valves in liquid service. The choice of the interval
affects the emission reduction achievable since more frequent inspection
will result in leaking sources being found and fixed sooner. In order
to evaluate the effectiveness of different inspection intervals, it is
-------
necessary to estimate the rate at which new leaks will occur and repaired
leaks will recur. The estimates which have been used to evaluate yearly,
quarterly, and monthly inspections are shown in Table 4-2.
4.1.3.3 Allowable Interval Before Repair. If a leak is detected,
the equipment should be repaired within a certain time period. The
allowable repair time should reflect an interest in eliminating a source
of VOC emissions but should also allow the plant operator sufficient time
to obtain necessary repair parts and maintain some degree of flexibility
in overall plant maintenance scheduling. The determination of this
allowable repair time will affect emission reductions by influencing the
length of time that leaking sources are allowed to continue to emit
pollutants. Some of the components with concentrations in excess of the
leak definition action level may not be able to be repaired until the
next scheduled unit shutdown, e.g., a unit turnaround.
The effects of different allowable repair intervals are shown in
Table 4-3. The percentages shown 1n the table are the percent of emis-
sions from the component which would be affected by the repair if all other
contributing factors were 100 percent efficient. The emissions which occur
between the time the leak is detected and repair is attempted are increased
with increasing allowable repair intervals.
4.1.3.4 Achievable Emission Reduction. Repair of leaking components
will not always result in complete emission reduction. The. repair of
components which have initial monitoring levels below 1,000 ppm has not
been adequately demonstrated. Repair of those components with low initial
leak rates may actually result in an emission rate increase. However, in
order to estimate repair effectiveness, it was assumed that emissions could
be reduced to a level of 1,000 ppm. The average emission rates of components
above 10,000 ppm and at 1,000 ppm are shown in Table 4-4.
4.1.3.5 Development of Controlled Emission Factors. The uncon-
trolled emission levels for the emission sources that are typically found
in the model plants were previously presented 1n Chapter 3 (Table 3-1).
Controlled VOC emission levels can be calculated by a "controlled emission"
factor. This factor can be developed for each type of emission source by
using the general expression:
4-9
-------
TABLE 4-2. ESTIMATED OCCURRENCE AND RECURRENCE RATE OF LEAKS FOR VARIOUS MONITORING INTERVALS
Estimated percer
of sources leaki
at above 10,000 ;
Source type Initially3
Pump seals
Light liquid service
Heavy liquid service
In-Tine valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Compressor seals
Flanges
23
2
10
12
0
8
33
0
Estimated percent of
)t initial leaks which
'n9 are found leaking at
'P"1 subsequent inspections b
Annual
20
20
20
20
20
20
20
20
quarterly
10
10
10
10
10
10
10
10
Monthly
5
5
5
5
5
5
5
5
Estimated percent of
sources which are
found leaking at
subsequent inspections c
Annual Quarterly
4.6
0.4
2.0
2.4 •'
0.0
1.6
6.3
0.0
2.3
0.2
1.0
1.2
0.0
0.8
3.3
0.0
Monthly
1.2
0.1
0.5
0.6
0.0
0.4
1.7
0.0
^Approximate fraction of sources having leaks equal to or greater than 10,000 ppm prior to repair.
Approximate fraction of leaking sources that were repaired but found to leak during subsequent
inspections. These approximations are based on engineering judgment.
Approximate fraction of sources that were repaired but found to leak during a subsequent inspection.
These approximations are the product of the information presented in footnotes a and b.
-------
TABLE 4-3. MAXIMUM POTENTIAL CONTF-'OL EFFICIENCY AS A FUNCTION OF REPAIR
INTERVAL ASSUMING 100 PERCENT EFFICIENCY FOR OTHER FACTORS3
Allowable repair interval (days)
30
15
Percent of emissions affected
95.9 97.9
99.3 99.9
Assumes that efficiencies of all other control factors (action level,
achievable emission reduction, monitoring frequency) are 100 percent.
TABLE 4-4. AVERAGE EMISSION RATES FROM SOURCES
ABOVE 10,000 PPMV AND AT 1000 PPMV'U
Source type
Pump seals
Light liquid service
Heavy liquid service
In-line valves
Vapor service
Light liquid service
Heavy 1 iquid service
Safety/relief valves
Compressor seals
Flanges
(V)
Emission rate
from sources above
10,000 ppmv3
(kg/hr)
0.45
0.21
0.21
0.07
0.005
1.4
l.'l
0.003
(X-)
Emission rate
from sources at
1000 ppmv
(kg/hr)
0.035
0.035
0.001
0.004
0.004
0.035
0.035
0.002
Percentage
reduction
92.0
83.0
99.5
94.0
20.0
97.5
97.0
33.0
... . ... . - ing
screening values above 10,000 ppmv.
Emission rate of all sources, within a source type, havina screening
values of 1000 ppmv.
4-11
-------
Controlled emission factor = Uncontrolled factor - uncontrolled
factor x emission reduction efficiency
The reduction efficiency can be developed by the following expression and
correction factors:
Reduction efficiency = A x B x C x D^1
Where:
A = Theoretical Maximum Control Efficiency = fraction of total mass
emissions for each source type with VOC concentrations greater
than the action level (Table 4-1, Figure 4-1).
B = Leak Occurrence and Recurrence Correction Factor = correction
factor to account for sources which start to leak between
inspections (occurrence) ; for sources which are found to
be leaking, are repaired and start to leak again before the
next inspection (recurrence) (Tables 4-2, 4-6); and for known leaks
which are not repaired.
C = Non-Instantaneous Repair Correction Factor = correction factor
to account for emissions which occur between detection of a leak
and subsequent repair; that is, repair is not instantaneous
(Table 4-3).
D = Imperfect Repair Correction Factor = correction factor to
account for the fact that some sources which are repaired are
not reduced to zero emission levels. For computational pur-
poses, all sources which are repaired are assumed to be reduced
, to a 1000 ppm emission level (Table 4-4).
These correction factors can, in turn, be determined from the following
expressions: —
0) B = 1 - /
(?) C = 365 " l
UJ L 365
(3) D = 1 - f
4-12
-------
Where:
n" = Average number of leaks occurring and recurring over the
monitoring interval (including known leaks which were not repaired).
N = Total number of sources at or above the action level (Figure
4-2).
t = Average time before repairs are made (with a 15-day repair limit,
7.5 is the average used).
f = Average emission factor for sources at the average screening
value achieved by repair.
F = Average emission factor for all sources at or above the action
level.
An example of a control effectiveness calculation is present.,"! in Tahlr: A-5.
Support data for this calculation are presented in Tables 4-1, 4-2, 4-3,
4-4, and 4-6, as well as in Figures 4-1 and 4-2.
4.2 EQUIPMENT SPECIFICATIONS
Fugitive emissions may be reduced by using process equipment which is
designed to prevent leakage. Equipment specifications for each emission source
are described below. Some of the specifications may be applicable to more
than one type of source.
4.2.1 Pumps
Fugitive emissions from pumps occur at the junction of a moving shaft
and a stationary casing. Equipment specifications that may be implemented
for pumps include elimination of this junction, improvement of the seal at
the junction, or collection and control of the emissions from the junction.
4.2.1.1 Seal!ess Pumps. Pumps such as diaphragm type pumps or "canned"
pumps do not have a shaft/casing junction and therefore do not leak the
pumped fluid in the normal course of operation. However, failure of the
diaphragm in a diaphragm pump may result in temporary emissions of VOC.
Seal less pumps are used primarily in SOCMI processes where the pumped fluid
is hazardous or toxic, and every effort must be made to prevent leaks of the
fluid.
/I-13
-------
TABLE 4-5. EXAMPLE OF CONTROL EFFICIENCY CALCULATION
Assume:
1) A leak detection and repair program to reduce emissions from
valves in gas/vapor source.
2) Action level = 10,000 ppm.
3) Average screening value after directed repair = 1,000 ppm.
4) Leak detection monitoring interval = 3 months.
5) Allowable repair interval = 15 days.
6) Number of valves having new or recurring leaks between repair
intervals, n = 0.2N (see Table 4-6).
m
Calculations:
A = 0.98 (from Figure 4-1 for a screening value of 10,000 ppmv)
B = 0.9 (from Table 4-6)
C = 0.979 (from Table 4-3 for 15-day interval)
where:
P _ A(Avg. uncontrolled emission factor)3
Fraction of sources screening > 10,000
= (0.98)(0.021 kg/hr)/0.10 = 0.206 kg/hr
f = Emission factor at 1000 ppmc
= 0,001 kg/hr
and 0 = (1 - 4 = 0.995
Overall percentage reduction =AxBxCxD
= (0.98) x (0.9) x (0.979) x (0.995)
= 86 Percent
Therefore:
Control effectiveness factor = 0.021 kg/hr - (0.86) (0.021 kg/hr)
= 0.003 kg/hr
? Reference 12. '
From Figure 4-2.
c Reference 13.
4-14
-------
TABLE 4-6. IMPACT OF MONITORING INTERVAL ON CORRECTION FACTOR ACCOUNTING
FOR LEAK OCCURRENCE/RECURRENCE (FOR EXAMPLE CALCULATION)
Monitoring a - b c
interval m _m B
1 month 0.1Nd 0.05N 0.95
3 months 0.2N 0.1N 0.90
1 year 0.4N 0.2N 0.80
a n. = Total number of leaks which occur, recur, and remain between
m monitoring intervals.
n = Average number of leaks over the monitoring interval.
c B = Correction factor accounting for leak occurrence/recurrence.
d N = Total number of sources at or above the action level.
4-15
-------
UPPER LIMIT OF 90%
CONFIDENCE INTERVAL
ESTIMATED PERCENT OF
TOTAL MASS EMISSIONS
LOWER LIMIT OF 90%
CONFIDENCE INTERVAL
PERCENT OF TOTAL MASS
EMISSIONS -
INDICATES THE PERCENT OF
TOTAL EMISSIONS ATTRIBU
TABLE TO SOURCES WITH
SCREENING VALUES GREATER
THAN THE SELECTED VALUE
10 100 1000 10.000 100.000 1,000,000
SCREENING VALUE (ppmv) (LOG10 SCALE)
Figure 4-1. Cumulative distribution of total emissions by screening
values - valves - gas/vapor streams.
111
O
-PERCENT OF SOURCES
PERCENT OF SOURCES -
INDICATES THE PERCENT OF
SOURCES WITH SCREENING
VALUES GREATER THAN THE
SELECTED SCREENING VALVE
LOWER LIMIT OF THE
„ 95% CONFIDENCE
INTERVAL
1 10 100 1000 10.000 100.000 1.000.000
SCREENING VALUE
-------
4.2.1.2 Dual Mechanical Seals. Dual mechanical seals consist of
two mechanical sealing elements usually arranged in a back-to-back or
tandem configuration. In both configurations a barrier fluid circulates
between the seals. In the back-to-back arrangement the barrier fluid
system is at a higher pressure than the pressure in the seal area.
Therefore, any leakage of barrier fluid would be across the inner seal
into the product and across the outer seal to the environment. In the
tandem configuration the barrier fluid may be at a lower pressure than
that of the seal area. If the pressure in the barrier fluid system is
lower, any leakage of product would occur across the inner seal into
the barrier fluid. Any leaks into the barrier fluid may be dissolved or
suspended in the barrier fluid, and subsequent degassing of the barrier
fluid may result in emissions of VOC. Therefore, barrier fluid degassing
vents would have to be controlled to provide maximum control effectiveness
of dual mechanical seals.
The barrier fluid system may be a circulating system or it may rely
on convection to circulate fluid within the system. While the barrier
fluid's main function is to keep the pumped fluid away from the environment,
it can serve other functions as well. A barrier fluid can provide
temperature control in the stuffing box. It can also protect the pump
seals from the atmosphere, as in the case of pumping easily oxidizeable
materials which form abrasive oxides or polymers upon exposure to air.
A wide variety of fluids can be used as barrier fluids. Some of the
more common ones which have been used are water (or steam), glycols,
methanol, oil, and heat transfer fluid. In cases in which product
contamination cannot be tolerated, it may also be possible to use clean
product, a product additive, or a product diluent.
Emissions of VOC from degassing vents can be controlled by a closed
vent system which consists of piping and, if necessary, flow inducing
devices to transport the degassing emissions to a control device such as
a process heater, or vapor recovery system. Control effectiveness of a
dual mechanical seal and closed vent system is dependent on the effectiveness
of the control device used and the frequency of seal failure. Failure
4-17
-------
of both the inner and outer seals can result in relatively large VOC
emissions at the seal area of the pump. Pressure monitoring of the
barrier fluid may be used in order to detect failure of the seals. In
addition, visual inspection of the seal area also can be effective for
detecting failure of the outer seals. Upon seal failure, the leaking
pump would have to be shut down for repair.
Dual mechanical seals are used in many SOCMI process applications;
however, there are some conditions that preclude the use of dual mechanical
seals. Their maximum service temperature is usually limited to less
than 260°C, and mechanical seals cannot always be used successfully on
pumps with reciprocating shaft motion.
4.2.1.3 Closed Vent Systems. The system described above for controlling
degassing vent emissions could also be applied to control emissions from the
seal area of pumps. This application would require the use of some type
of flow inducing device to transport the emissions from the seal area to the
control device. The seal area would be enclosed in order to collect the
emissions and a vacuum eductor or a compressor could be used to remove vapors
from the seal area. However, normal pump operating practices may require
frequent visual inspection or mechanical adjustments in the seal area. This
would not be possible with a closed vent system at the seal area. A potential
problem with this approach is that explosive mixtures may be created by
enclosing the pump seal area, and therefore safety and operating practices
may limit the use of closed vent systems for pump seal areas.
4.2.1.4 Control Device. Several types of controls could be used to
dispose of VOC emissions trapped in the pump seal barrier fluid. Incineration,
carbon adsorption, and condensation are three control methods which might
typically be applied. Control efficiencies of the three methods are dependent
on specific operating characteristics and types of VOC. However, incineration
can achieve bettor than 95 percent efficiency.17 Temperature and residence
time affect the VOC destruction efficiency. A temperature of 1400°F and a
residence time of 0.5 seconds residence time results in > 90 percent efficiency.
A temperature of 1500°F combined with j residence time of 0.5 seconds gives
>_98 percent VOC destruction.18
4-18
-------
Carbon adsorption systems can achieve 95-99 percent control efficiency
through proper design and operation.^ Condensation systems can achieve
>90 percent VOC capture.20
Flares, while they are commonly used in chemical plants, would not be as
applicable to these small vent streams as they are to larger streams. Flare
efficiency can vary from 60 to 99 percent?1 depending on how closely the
design specifications match the flow characteristics of the VOC conveyed to
the flare.
4.2.2 Compressors
Fugitive emissions from compressors occur at the junction of a moving
shaft and a stationary casing. Emission reductions from this source type
may be achieved by improving the seal at the junction, or collecting and
controlling the emissions from the junction.
4.2.2.1 Mechanical Contact. Mechanical contact seals for compressors
are similar to the mechanical seals described for pump applications. However,
compressors in some services cannot be fitted with mechanical contact
seals. Existing compressors may have mechanical contact seals equipped
with seal oil flush systems. Seal oil reservoir degassing vents must be
controlled with closed vent systems as described for pumps. Sometimes a
buffer or barrier gas may be used to form a buffer between the compressed
gas and the atmosphere. This system requires a clean external gas
supply which is compatible with the qas bein
-------
4.2.2.3 Control Device. Several types of controls could be used to
dispose of VOC emissions collected from compressor seal areas. Incineration,
carbon adsorption, and condensation are three control methods which might
typically be applied. Control efficiencies of the three methods are dependent
on specific operating characteristics and types of VOC. However, incineration
can achieve better than 95 percent efficiency.^ Temperature and residence
time affect the VOC destruction efficiency. A temperature of 1400°F and a
residence time of 0.5 seconds residence time results in 90 percent efficiency.
A temperature of 1500°F combined with a residence of 0.5 seconds gives >98
percent VOC destruction.23
Carbon adsorption systems can achieve 95-99 percent control efficiency
through proper design and operation/-^ Condensation systems can achieve >90
percent VOC capture/-^
Flares, while they are commonly used in chemical plants, would not be
as applicable to these small vent streams as they are to larger streams.
Flare efficiency can vary from 60 to 99 percent?6 depending on how closely
the design specifications match the flow characteristics of the VOC conveyed
to the flare.
4.2.3 Pressure Relief Devices
Pressure relief devices include rupture disks and safety/relief valves.
Fugitive emissions from these devices occur because of improper seating or
partial failure of the device. These fugitive emissions do not include
emissions which result from normal operation of the devices caused by over-
pressure of the process or vessel which the device protects. Fugitive
emissions from rupture disks may be caused by pinhole leaks in the disk
itself caused by corrosion or fatigue. Fugitive emissions from relief valves
may be caused by failure of the valve seating surfaces, improper reseating
after overpressure relieving, or process operation near the relief valve
set pressure which may cause "simmerinq".
4.2.3.1 Rupture Disks. Although they are also pressure relief devices,
rupture disks can be installed upstream of a safety/relief valve in order
to prevent fugitive emissions through the relief valve seat. This procedure
may require use of a larger size relief valve because of operating codes. The
14-ZO
-------
disk/valve combination may also require appropriate piping changes to prevent
disk fragments from lodging in and damaging the relief valve when relieving
overpressure. A block valve upstream of the rupture disk is also required
in order to permit in-service replacement of the disk after overpressuring.
If the disk could not be replaced, the first overpressure would result in
the relief valve being the same as an uncontrolled relief valve. In some
chemical plants, installation of a block valve upstream of a pressure
relief device may be a common practice. While it is allowed by ASME codes,
it may be forbidden by operating or safety procedures for a particular
company. Tandem pressure relief devices with a three-way valve can be used
to avoid operation without overpressure protection. Rupture disk/relief
valve combinations must have some provision for testing the integrity of
the disk. The area between the rupture disk and relief valve must be
connected to a pressure indicator, recorder, or alarm. If the process fluid
is not hazardous or toxic, a simple bubbler apparatus could be used to
test disk integrity by connecting the bubbler to the disk/valve area. -.The-
control efficiency of the disk valve combination is assumed to be 100 percent
for fugitive emissions. If the disk integrity is not maintained or if the
disk is not replaced after overpressure relief, the control efficiency would
be lowered. The disk/valve combination has no effect on emissions which
result from overpressure relieving.
4.2.3.2 Resilient Seat Relief Valves. Manufacturers of relief valves
state that resilient seat or "0-ring" relief valves provide better reseat
qualities compared to standard relief valves. No test data are available to
verify these statements. These improvements would have no effect on over-
pressure emissions or fugitive emissions due to seal failure or "simmering".
4.2.3.3 Closed Vent Systems. A closed vent system can be used to
transport the discharge or leakage of pressure relief devices to a control
device such as a flare. Since overpressure discharges as well as fugitive
emissions are routed to the control device, it must be sized appropriately.
A larger pressure relief device may be required for use with a closed vent
system. The control efficiency of a closed vent system is dependent on the
effectiveness of the control device. Typical flare systems may be only
4-21
-------
60 percent effective for fugitive emission destruction.28 This efficiency
reflects the fact that many flare systems are not of optimum design. Flares
that are designed to handle large volumes of vapors associated with over-
pressure releases may also be used to handle low volumes of fugitive emissions.
With such designs, optimum mixing is not achieved because" the"vent gas exit
velocity is low and large flares generally cannot properly inject steam into
low volume streams.29 A properly designed flare system typically exhibits a
99 percent hydrocarbon destruction efficiency.30 Closed vent systems for
pressure relief devices are used in existing SOCMI processes especially
where the emissions may be hazardous or toxic.
4.2.4 Open-Ended Valves
Fugitive emissions from open-ended valves are caused by leakage through
the seat of the valve. Emissions may also occur through the stem and gland
of the valve, and these emissions may be controlled by methods described
for valves in Section 4.1.2. Approximately 28 percent of SOCMI valves
(excluding safety/relief and check valves) in VOC service are open-ended.
They include drain, purge, sample, and vent valves. Fugitive emissions from
open-ended valves can be controlled by installing a cap, plug, flange, or
second valve to the open end of the ..valve. In the case of a second valve, the
upstream valve should always be closed first after use of the valves. Each
time the cap, plug, flange, or second valve is opened, any VOC which has
leaked through the first valve seat will be released. These emissions have not
been quantified. The control efficiency will be dependent on the freaucncy of
removal of the cap or plug. Caps, plugs, etc. for open-ended valves do not
affect emissions which may occur during use of the valve. These emissions may
be caused by line purging for sampling, draining or venting through the
open-ended valve. Caps, plugs, flanges, or second valves for open-ended
valves are required by California regulations.-^ _..
4.2.5 Sampjjncj Connectiojis
liKlilive emissions from sninpl in«i connect, ions occur .is o result of
pm'<|1ii<| I he Siimp'l Iru] line in order1 l,.o oM.iiri .1 represent! I i ve s.niiple of t:he.
process fluid. Approximately ^5 percent of opon-endo
-------
designed so that the purged fluid is returned to the process at a point of
lower pressure. A throttle valve or other device is required to induce the
pressure drop across the sample loop. Closed loop sampling is assumed to
be 100 percent effective for controlling fugitive emissions. The purged
fluid could also be directed to a control device such as a flare. In this
case the control efficiency would be dependent on the flare efficiency for
hydrocarbon destruction. Since some pressure drop is required to purge
sample through the loop, low pressure processes, or tankage may not be
amenable to closed loop samplinq. Safety requirements may prohibit closed
loop sampling in some instances.
4.2.6 In-Line Valves
Fugitive emissions from valves occur at the stem or gland area of the
valve body. Diaphragm and bellows seal valves do not have a stem or gland
and therefore are not prone to fugitive emissions. They are generally used
where hazardous or toxic process fluids are present and fugitive emissions
must be eliminated. Their control effectiveness is approximately 100 percent,
although a failure of the diaphragm or bellows may cause large temporary
emissions. The applicability of these types of valves is limited. They may
not be suitable for many applications because of process conditions or cost
consideration.
4.2.7 Effectiveness of Equipment Specifications
In order to quantify the environmental and economic impacts of applying •-
controls, the control efficiency must be determined. In some cases, there
are many complicating factors which must be considered in estimating control
efficiency. For example, the efficiency of caps or plugs for open-ended
valves is dependent on 1) the frequency of removal of the cap or plug, and
2) the emission rate through the valve seat. Estimated control efficiencies
for various equipment modifications are shown in Table 4-7. These estimates
represent the maximum emission reduction possible for the equipment modifi-
cations. In some instances, the actual emission reduction will depend on
other factors such as the efficiency of control devices attached to closed
vent systems. Carbon absorption or vapor recovery systems would approach
100 percent efficiency, but flares may be only 60 percent effective for
hydrocarbon destruction. The estimates of effectiveness shown in Table 4-7
'were used to calculate environmental tind economic impacts of regulatory
alternatives in Chapters 7 and 8 of this document.
4-23
-------
TABLE 4-7. EFFECTIVENESS OF EQUIPMENT MODIFICATIONS
Source type/ Control efficiency
equipment modification (%]_
Pumps
Sealless pumps 100
Double mechanical seals/closed vent system
Closed vent system on seal area
Compressors
Double mechanical seals/closed vent system VIOOa
Closed vent system on seal area %100a
Safety/relief valves
Closed vent system 60
Rupture disks 100
Open-ended lines
Caps, plugs, blinds, second valves 100 c
Sampling connections
Closed loop sampling 100
In-line valves
Diaphragm valves 100
Bellows-sealed valves 100
aAlthough a control efficiency is not attained in all cases, it is
achievable in some cases.
This control effectiveness reflects the fact that a closed vent system is
normally sized for emergency relief.
cThis control efficiency reflects the use of these devices downstream of
an initial valve with VOC on one side and atmosphere on the other.
4-24
-------
4.3 REFERENCES
1. Hustvedt, K. C. and R. C. Weber. (U. S. Environmental Protection
Agency.) Detection of Volatile Organic Compound Emissions from
Equipment Leaks. (Presented at the 71st Annual Meeting of the Air
Pollution Control Association. Houston. June 25-30, 1978.) 8 p.
2. Reference 1.
3. Teller, J. H. Advantages Found in On-Line Leak Sealing. The Oil
and Gas Journal. 77(29):54-59. July 16, 1979.
4. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979. p. 2.
5. Reference 4.
6. Reference 4.
7. Letter and attachments from Bottomley, F. R., Union Oil Company, to
Feldstein, M., Bay Area Air Quality Management District. April 10,
1979. 33 p. Information about valve repairability.
8. U. S. Environmental Protection Agency. Control of Volatile Organic
Compound Leaks from Petroleum Refinery Equipment. Research Triangle
Park, N. C. Publication No. EPA-450/2-78-036. June 1978.
9. Reference 4.
10. Reference 4.
11. Rosebrook, D. D. (Radian Corporation.) Proceedings: Symposium on
Atmospheric Emissions from Petroleum Refineries. (Prepared for
U. S. Environmental Protection Agency.) Research Triangle Park,
N. C. Publication No. EPA-600/9-80-013. March 1980. pp. 421-440.
12. Reference 4.
13. Reference 4.
14. Reference 4.
15. Reference 4.
4-25
-------
16. Erikson, D. G. and V. Kalcevic. (Hydroscience, Inc.) Emissions
Control Options for the Synthetic Organic Chemicals Manufacturing
Industry. (Prepared for L). S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577. February 1979.
p. III-l.
17. Radian Corporation. Control Techniques for Volatile Organic Emissions
from Stationary Sources. (Prepared for U. S. Environmental Protection
Agency.) Research Triangle Park, N. C. Publication No. EPA-450/2-
78-022. May 1978. p. 34.
18. Blackburn, J. W. (Hydroscience, Inc.) Emissions Control Options
for the Synthetic Organic Chemicals Manufacturing Industry - Thermal
Oxidation. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577. December 1979.
p. II-6.
19. Basdekis, H. S. (Hydroscience, Inc.) Emissions Control Options
for the Synthetic Organic Chemicals Manufacturing Industry - Carbon
Adsorption. (Prepared for L). S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577. February 1980.
p. 11-24.
20. Burklin, C. E., et al. (Radian Corporation.) Control of Hydrocarbon
Emissions from Petroleum Liquids. (Prepared for U. S. Environmental
Protection Agency.) Research Triangle Park, N. C. Publication No.
EPA-600/2-75-042. September 1975. p. 16.
21. L). S. Environmental Protection Agency. Draft Background Information
for Proposed Standards for Benzene Emissions from the Ethylbenzene/Styrene
Industry. Research Triangle Park, N. C. Publication No. EPA-
450/3-79-035a. October 1979.
22. Reference 17.
23. Reference 18.
24. Reference 19.
25. Reference 20.
26. Reference 21.
27. Part UG-General Requirements (Section VIII, Division I.) In: ASME
Boiler and Pressure Vessel Code, An American National Standard.
New York, The American Society of Mechanical Engineers, 1977.
p. 449.
28. Reference 21.
4-26
-------
29. Reference 17.
30. U. S. Environmental Protection Agency. Control of Volatile Organic
Emissions from Existing Stationary Sources, Volume I: Control
Methods for Surface-Coating Operations. Research Triangle Park,
N. C. Publication No. EPA-450/2-76-028. November 1976. p. 42.
31. Reference 16, p. III-5.
32. Reference 16.
33. Reference 17.
4-27
-------
-------
5. MODIFICATION AND RECONSTRUCTION
In accordance with the provisions of the Code of Federal Regulation
Title 40, Sections 60.14 and 60.15 (40 CFR 60.14 and 60.15), an "existing
facility" can become an affected facility and, subsequently, subject to
the standards of performance if it is modified or reconstructed. An
existing facility, as defined in 40 CFR 60.2, is a facility of the type
for which standards of performance have been promulgated and the construction
or modification of which was begun prior to the proposal'date of the
applicable standards.
The applicability of provisions 40 CFR 60.14 and 60.15 to the SOCMI,
and the conditions, as outlined in these provisions, under which existing
facilities could become subject to standards of performance are discussed
below.
5.1 GENERAL DISCUSSION OF MODIFICATION AND RECONSTRUCTION PROVISIONS
5.1.1 Modification
"Modification" is defined in 40 CFR 60.14 (a) as any physical or
operational change of an existing facility which increases the emission rate
of any pollutant to which a standard applies. Exceptions to this definition
are presented in paragraphs (e) and (f) of Section 60.14. These exceptions
are as follows:
Paragraph (e) - Physical or operational changes to an existing
facility which will not be considered modifications are
specified in this portion of Section 60.14. These changes
include:
a. Routine maintenance, repair, and replacement.
b. An increase in the production rate not requiring
a capital expenditure as defined in Section
60.2 (bb).
5-1
-------
c. An increase in the hours of operation.
d. Use of an alternative fuel or raw material if
prior to the standard the existing facility
was designed to accommodate that alternate fuel
or raw material.
e. The addition or use of any system or device
whose primary function is the reduction of
air pollutants, except when an emission control
system is removed or replaced by a system con-
sidered to be less efficient.
f. Relocation or change in ownership.
Paragraph (f) - This paragraph provides for superceding
any conflicting provisions of this section.
Upon modification, an existing facility becomes an affected facility for
each pollutant to which a standard applies and for which there is an increase
in the emission rate to the atmosphere.
5.1.2 Reconstruction
Under the provisions of Section 60.15, an existing facility becomes
an affected facility upon reconstruction, irrespective of any change in
emission rate. Generally, reconstruction is considered to occur upon the
replacement of components if the fixed capital cost of the new components
exceeds 50 percent of the fixed capital cost that would be required to
construct a comparable entirely new facility, and it is economically and
technically feasible for the facility to comply with the applicable standards
of performance. The final judgments on what replacement constitutes recon-
struction and when it is technologically and economically feasible to comply
with the applicable standards of performance is made by the Administrator.
The Administrator's final determinations are made on the following bases:
(1) comparison of the fixed capital costs of the replacement
components and a newly constructed comparable facility,
(2) the estimated life of the facility after the replacements
compared to the life of a comparable entirely new facility,
(3) the extent to which the components being replaced cause or
contribute to the emissions from the facility, and
5-2
-------
(4) any economic or technical limitations on compliance with
applicable standards of performance which are inherent in
the proposed replacements.
The purpose of this provision is to ensure that an owner or operator
does not perpetuate an existing facility by replacing all but vestigial
components, support structures, frames, housing, etc., rather than
totally replacing it in order to avoid subjugation to applicable standards
of performance. In accordance with Section 60.5, EPA will, upon request,
determine if the action taken constitutes construction (including recon-
struction) .
5.2 APPLICABILITY OF MODIFICATION AND RECONSTRUCTION PROVISIONS TO THE
SOCMI
5.2.1 Modification
Changes in operating conditions would mean that a facility would be
subject to new source standards of performance if the changes made cause
increased emissions. Under these conditions the facility becomes a modified
facility. Several changes in operating conditions that could be encountered
in an organic chemical plant are presented below. The possible effects of
these changes on emissions are presented.
Routine changes and additions of fugitive emission sources are
commonly made to increase ease of maintenance, to increase productivity,
to improve plant safety, and to correct minor design flaws. These
additions of fugitive emission sources would cause an increase in fugitive
emissions. However, fugitive emissions from other sources could be
reduced to compensate for this increase.
The replacement of a potential fugitive emission source such as a pump
or valve commonly occurs in an organic chemical plant. If such a source
is replaced with an equivalent source (such as is done during routine
repair and replacement), the fugitive emissions from the facility should not
increase because the number of potential sources in the same vapor pressure
service (handling the same organic chemical) remains unchanged.
5-3
-------
Process equipment pieces such as heat exchangers, reactors, distillation
columns, reboilers, filters and separators, or new control loops are
commonly added to existing facilities in the organic chemical industry
to increase the capacity of or to optimize a process. The addition of
this equipment would normally increase fugitive emissions from a facility
due to the increased number of potential emission sources (pumps, valves,
sampling connections, etc.) that are associated with the process equipment.
In some cases a facility in the organic ehcmical industry can be
converted from the production of one chemical to the production of a
second chemical. This normally occurs when production of the second
chemical results in greater profits. In such a case, whenever either
the number of fugitive emission sources or the vapor pressure of the
second chemical increases during this conversion, the level of VOC
emissions from the facility could be expected to increase. As shown in
Table 3-1, emission factors for equipment in vapor service are higher
than emission factors for equipment in light liquid service which are
higher than emission factors for equipment in heavy liquid service. So
that, if the vapor pressure of the second chemical is higher than the
vapor pressure of the first chemical, the fugitive emissions could be
expected to increase.
Changes may be made to a process, although the chemical being
produced remains the same. One such case would be a change in catalyst
for producing a given chemical. In such a case the level of fugitive
emissions would not be expected to change because neither the number of
sources nor the vapor pressure of the chemical would change.
In many cases, there may be a desire to increase the capacity of an
existing facility. This may be achieved by replacing certain process
equipment (pumps, heat exchangers, reactors, etc.) with similar equipment
but of larger capacity or addition or process equipment. If this replacement
or addition does not increase the number of fugitive emission sources
handling the given organic chemical, the level of fugitive emissions
would not be expected to increase. However, if the number of sources
were to increase due to this replacement or addition, then VOC emissions
could be expected to increase.
5-4
-------
5.2.2 Reconstruction
When an owner or operator replaces several components of an existing
facility, that facility may become subject to applicable standards of
performance under the provisions of Section 60.15. For example, if an
owner or operator replaces several fugitive emission sources such as
pumps, compressors, or sampling loops in an existing facility, and if
the fixed capital costs for the new equipment exceeds 50 percent of the
costs of all fugitive emissions sources in the unit, the Administrator
may determine that reconstruction has occured. Reconstructions may
occur as a result of damage caused by fires, explosions, hurricanes, or
other catastrophes. They might also result, from feedstock changes,
product changes, or other major process changes which would require
additions or replacement of several fugitive emission sources.
5-5
-------
-------
6. MODEL PROCESS UNITS AND REGULATORY ALTERNATIVES
This chapter presents model process unit parameters and alternative
emission controls considered for reduction of fugitive emissions from SOCMI
sources. The model units were selected to represent the range of processing
complexity in the industry. They provide a basis for comparing environmental
and economic impacts of the regulatory alternatives. The regulatory alter-
natives selected provide varying levels of emission control.
6.1 MODEL UNITS
Available data show that fugitive emissions are proportional to the
number of potential sources, but are not related to capacity, throughput,
age, temperature, or pressure. Therefore SOCMI model units defined for this
analysis represent different levels of process complexity (number of sources)
rather than different unit sizes.
6.1.1 Sources of Fugitive Emissions
The various potential fugitive emission sources in a SOCMI process
unit were described in Chapter 3. Data from petroleum refineries indicate
o
that cooling towers are very small sources of VOC emissions. Differences
in SOCMI operating procedures, such as recirculation of process water, might
result in cooling tower VOC emissions, but no data are available to verify
this. The number of agitator seals in SOCMI is not known. Furthermore, the
emission rate from SOCMI agitator seals has not been measured. Since there
are no data from similar sources in other industries, no estimates of emission
rate can be made. Because of these uncertainties, cooling towers and agitator
seals are not included in the Model Units.
6-1
-------
6.1.2 Model Unit Parameters
In order to estimate emissions, control costs, and environmental impacts
for SOCMI units on a unit specific basis, three model units were developed.
The technical parameters for the model units are shown in Table 6-1. These
three model units represent the range of emission source populations that may
exist in SOCMI process units. The technical parameters were developed from
o
a data base compiled by Hydroscience, Inc. The data base included equipment
source counts from 62 SOCMI plants which produce 35 different chemicals.
These plant sites represent approximately 5 percent of the total existing
SOCMI plants and include large and small capacities, batch and continuous
production methods, and varying levels of process complexity. The source
counts for the 35 chemicals include pumps, valves, and compressors. These
counts were used in combination with the number of sites which produce
each chemical in order to determine the average number of sources per site.
Hydroscience estimates that 52 percent of existing SOCMI plants are similar
to Model Unit A, 33 percent are similar to B, and 15 percent are similar to C.
Data from petroleum refineries indicate that emission rates of sources
decrease as the vapor pressure (volatility) of the process fluid decreases.
Three classes of volatility have been established based on the petroleum
refinery data. These include gas/vapor service, light liquid service, and
heavy liquid service.^ The split between light and heavy liquids for the
refinery data is between streams called naphtha and kerosene. Since simi-
lar stream names may have different vapor pressures, depending on site
specific factors, it is difficult to quantify the light-heavy split. The
break point is approximately at a vapor pressure of 0.3 kPa at 20°C.
The data collected by Hydroscience were used to estimate the split between
gas/vapor and liquid service for each source type. In order to apply
emission factors for liqht and heavy liquid service, it is assumed that
one half of SOCMI liquid service sources are in liqht liquid service. There
are no data available on the actual distribution of" sources in volatility
ranges. It is assumed that all SOCMI packed seal pumps are in heavy liquid
service. This assumption is reasonable, since more volatile liquids are
6-2
-------
TABLE 6-1. FUGITIVE EMISSION SOURCES FOR THREE MODEL UNITS'
Number of components in model unit
Equipment component3
Pump seals
Light 1 iquid service
Single mechanical
Dual mechanical
Sealless
Heavy liquid service
Single mechanical
Packed
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Vapor service
Light liquid service
Heavy liquid service
Open-ended valves 'and linesc
Vapor service
Light liquid service
Heavy liquid service
Compressor seals
Sampling connections
Flanges
Cooling towers
Model unit
A
5
3.
0
5
2
90
84
84
11
1
1
9
47
48
1
26
600
Model unit
B
19
10
1
24
6
365
335
335
42
4
4
37
189
189
2
104
2400
Model unit
C
60
31
1
73
20
1117
1037
1037
130
13
14
115
581
581
8
320
7400
^Equipment components in VOC service only.
52% of existing units are similar to Model Unit A.
33% of existing units are similar to Model Unit B.
15% of existing units are similar to Model Unit C.
^Sample, drain, purge valves and the associated open end.
Based on 25% of open-ended valves. From Ref. 3, pg. IV-3,
eData not available.
6-3
-------
more suitable for mechanical seal applications, and newer process units tend
to use fewer packed seals. Sampling connections are a subset of the open-
ended valve category. Approximately 25 percent of open-ended valves are used
for sampling connections.7 Emissions which occur through the valve stem,
gland, and open-end are included in the open-ended valve category. The
emission factor for sampling connections applies only to emissions which
result from sample purging.
6.2 REGULATORY ALTERNATIVES
The purpose of developing different regulatory alternatives is to
provide a basis for determining the air-quality and non air-quality environ-
mental impacts, energy requirements, and the costs associated with varying
degrees of VOC fugitive emissions reduction. Regulatory alternatives represent
comprehensive programs for reduction of emissions. They are constructed by
making different combinations of control techniques described in Chapter 4.
The regulatory alternatives selected for analysis include a "status quo
of fugitive emission control" case and three increasingly restrictive levels
of emission control requirements. The "status quo" case allows for the
analysis of not implementing standards of performance. The three increasingly
restrictive control requirements allow for analysis of the impacts of different
systems with varying degrees of emission reduction. The requirements for
each of these regulatory alternatives .ire summarized in Table 6-? and are
described below.
6.2. I Keriu 1
-------
TABLE 6-2. REGULATORY ALTERNATIVES FOR FUGITIVE EMISSION SOURCES IN SOCMI
Source typea
Pumps
Liqht liquids
with single mechanical seals
with double mechanical seals
with no seals
Heavy liquids
with packed seals
with single mechanical seals
Valves (in-line)
Gas
Light liquid
CP, Heavy liquid
i
<-" Safety/relief valves
Gas
Light liquid
Heavy liquid
Open-ended valves and lines
Gas
Light liquid
Heavy liquid
Flanges
Sampling connections
Compressor seals
Monitoring
interval
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
I
Equipment
specification
None
None
None
None
None
Now
None
None
None
None
Nc-ie
None
None
None
None
None
None
]
Monitoring
interval
Annual lyb
Annual lyb
None
None
Hone
Quarterly
Annually
None
Quarterlyc
None
None
Quarterly
Annually
None
None
Noned
Quarterly
Regulatory
n
Equipment
specification
None
'tone
None
None
None
v
"(one
None
None
None
None
None
Cap/
Capsf
Capsf
None
None
None
al ternative
Monitoring
interval
Monthlyb
Monthlyb
None
None
None
Monthly
Monthly
None
Honthly1-
None
None
Monthly
Monthly
None
None
Noned
Monthly
III
Equipment
specification
None
None
None
None
None
None
None
None
Hone
None
Ncr.e
Capsf
Capsf
Capsf
Hone
Hone
None
Monitoring
interval
Noneh
Noncb
None
None
None
Monthly
Monthly
None
Nonec
None
None
Monthly
Monthly
None
None
Noned
None
IV
Equipment
specification
Double seals;
degassing vents
connected to
control device8
Degassing vents
connected to
control device6
None
None
None
None
None
None
Upstream
rupture disks
None
None
Capsf
Capsf
Capsf
None
Closed loop
sampling
Seal area or
degassing vents
connected to
control device
.Sources In VOC service.
^Plus weekly visual inspection. If liquid leak is observed, instrument monitoring is required to determine if action level is being exceeded.
Homering IT, required after each ove^ i.ressure release. If it is found to be leaking, the valve will be repaired.
Included in open-ended valves.
eSea 11 ess pumps stay also be used.
fOr blinds, plugs, second valves.
- r
-------
6.2.2 Regulatory Alternative II
This alternative would require leak detection and repair methods as
in the petroleum refinery control techniques guideline (CTG), EPA-450/2-78-036.
Leak detection would be accomplished by checking equipment components for
emissions of VOC using a portable VOC detection instrument to sample and
analyze the air in close proximity to the potential leak area. A measured
VOC concentration greater than some predetermined level, known as an "action
level", would be defined as a leak that would require equipment repair. A
measured VOC concentration less than the action level would not require equip-
ment repair. The action level is defined as 10,000 ppmv VOC concentration
for all cases.
Quarterly monitoring of compressors, gas service relief valves, inline
valves, and open-ended valves would be required. Annual monitoring of light
liquid service pumps and valves would be required. Weekly visual inspections
of light liquid pump seals would also be required. Leaks detected visually
would require instrument monitoring to determine if the action level is
exceeded. Relief valve monitoring after over pressure relieving would be
required. Open-ended valves would be required to be sealed with a cap, blind,
plug, or another valve.
6.2.3 Regulatory Alternative III
Regulatory Alternative III would provide for more restrictive control
than Alternative II by increasing the inspections for all applicable equipment
to monthly. Increasing the inspection', would result in a reduction of
emissions from residual leaking source'.; i.e., those sources which are found
leaking and are repaired and recur before the next inspection and those
sources that begin leaking between inspection. Thus, although this alterna-
tive is similar in approach to'Alternative II, it provides for more emissions
reduction. The requirements for weekly visual pump seal inspections, relief
valve monitoring after over pressure, <>nd caps for open-ended valves are
similar to those for Alternative II.
6.2.4 UeguJ ajtory Al_tern_ati vo_ IV
Alternative IV would require equipment specifications instead of more
frequent equipment inspections. This
-------
Alternatives II and III. Closed loop sampling techniques would be
required and rupture disks would be required on gas service relief
valves venting to the atmosphere. Maintenance of the integrity of the
disk would be required and replacement of the disk would be required if
a failure were detected. No monitoring would be required for relief
valves which have rupture disks upstream or which vent to a control
device header. Compressor seal areas or degassing vents from seal oil
reservoirs, or both, would be required to be connected to a control
device with a closed vent system. Pumps in light liquid service would
be required to have dual mechanical seals with a barrier fluid system.
Degassing vents from the barrier fluid system would be required to be
connected to a control device with a closed vent system.
6.3 REFERENCES
1. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979. pp. 11-49.
2. Radian Corporation. Assessment of Atmospheric Emissions from
Petroleum Refining, Appendix B: Detailed Results. (Prepared for
U. S. Environmental Protection Agency.) Research Triangle Park,
N. C. Publication No. EPA-600/2-80-075c. April 1980. pp. 300-321.
3. Erikson, D. G. and V. Kalcevic. (Hydroscience, Inc.) Emissions
Control Options for the Synthetic Organic Chemicals Manufacturing
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577 February 1979.
pp. IV-1, IV-2.
4. Reference 3, p. II-9-13.
5. Reference 1, pp. 11-23.
6. Reference 3, p. 11-10.
7. Reference 3, p. IV-8.
6-7
-------
-------
7. ENVIRONMENTAL IMPACT
The environmental impacts that would result from implementing the
regulatory alternatives being considered in this study are examined in this
chapter. Included in this chapter are estimates of the controlled VOC
fugitive emissions and the incremental reductions in uncontrolled VOC emissions
that could be achieved under each of the alternatives. Also, the impacts of
these regulatory alternatives on water quality, waste water generation and
treatment, solid waste generation and treatment or disposal, and energy
consumption or savings are discussed.
7.1 IMPACT ON ATMOSPHERIC EMISSIONS
Implementation of Regulatory Alternatives II, III, or IV, would reduce
VOC fugitive emissions from the SOCMI. To quantify reductions, the controlled
VOC emission levels from emission sources in the model units (described in
Chapter 6) were estimated for each alternative. These emission levels are
presented below for individual emission sources, for model units in SOCMI,
and then for SOCMI as a whole.
7.1.1 Emission Source Characterization
As indicated in Chapter 6, a SOCMI model unit typically consists of
several types of process equipment that contribute to fugitive VOC emis-
sions. Under Regulatory Alternative I (baseline case), all these sources
are "uncontrolled" emission sources. However, if Regulatory Alternative
II, III, or IV were implemented, the emissions from some uncontrolled sources
would be reduced; these sources would subsequently become "controlled"
sources. Both the controlled and uncontrolled sources are important because
the total fugitive VOC emissions from the model units and ultimately the
SOCMI are the sum of emissions from both types of sources.
7-1
-------
7.1.2 Development of VOC Emission Levels,
The uncontrolled emission levels were previously presented in Chapter
3 (Table 3-1). Controlled emission levels were developed for those
sources that would be controlled by the implementation of a regulatory
alternative. These controlled fugitive emission levels were calculated by
multiplying the uncontrolled emissions from this equipment by a "control
efficiency" presented in Chapter 4, Tables 4-2 through 4-4. The resulting
controlled VOC emission factors for each source are presented in Tables
7-1, 7-2, and 7-3 for Regulatory Alternatives II, III, and IV, respectively.
To arrive at the controlled VOC emission factors, the total VOC fugitive
emissions from Model A, Model B, and Model C units in the SOCMI were deter-
mined under each regulatory alternative. Initially, emissions from each
source type within a model unit were estimated by using the model unit equip-
ment inventories presented in Table 6-1 and the source emission factors
presented in Tables 7-1, 7-2, and 7-3. These emissions estimates were then
used to estimate the VOC fugitive emissions from each of the three model units.
An example calculation is presented in Table 7-4 to illustrate the procedure
used. The example is an estimate of the total VOC fugitive emissions from a
model unit under Regulatory Alternative II. The total VOC fugitive emissions
calculated for the respective model units under each regulatory alternative
are presented in Table 7-5. Also presented in this table are the average
reductions (expressed in percentages) in the baseline emission levels that
result from implementing Regulatory Alternatives II, III, or IV. Incremental
reductions in fugitive emission levels achieved by implementing the alterna-
tives are also presented in Table 7-5.
7.1.3 Future Impact on VOC Fugitive Emissions
In order to assess the future impacts of the various regulatory
alternatives on VOC fugitive emissions from the SOCMI, the levels of
these emissions were estimated for a period of five years after implementation
of a regulatory alternative. These emissions were estimated by using:
1) the emission factors presented in Tables 7-1, 7-2,
and 7-3;
2) the industry population for the assumed base year
of 1980;
7-2
-------
TABLE 7-1. EMISSION FACTORS FOR SOURCES CONTROLLED UNDER REGULATORY ALTERNATIVE II '
I
00
Uncontrolled
emission source
Inspection3
interval
Uncontrolled
emission
TUL LUi ,
kg/hr Ac
Correction
factors
Bd C6 Df
Control
efficiency
(AxBxCxD)
Controlled9
emission
factor,
kg/hr
Pumps
Light liquid service Yearly
0.120 0.87 0.80 0.98 0.92
0.63
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Quarterly
Yearly
Quarterly
Quarterly
0.
0.
0.
0.
021
010
160
440
0.98
0.84
0.69
0.84
0.90
0.80
0.90
0.90
0.98
0.98
0.98
0.98
0.99
0.94
0.97
0.97
0.
0.
0.
0.
86
62
59
72
0
0
0
0
.003
.004
.067
.126
aFrom Table 6-2.
bFrom Table 3-1.
Theoretical maximum control efficiency.'
Leak occurrence and reoccurrence correction factor - assumed to be 0.80 for yearly inspection, 0.90
for quarterly inspection, and 0.95 for monthly inspection.
eNon-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
average repair time yields a 0.98 yearly correction factor [365 - (15/2)] -=- 365.
Imperfect repair correction factor - calculated as 1 - (f v F). Where f = average emission rate for
sources at 1000 ppm and F = average rate for emission sources greater than 10,000 ppm. '
^Controlled emission factor = uncontrolled emission factor x [1 - (A x B x C x D)].
-------
TABLE 7-2. EMISSION FACTORS FOR SOURCES CONTROLLED UNDER REGULATORY ALTERNATIVE III
Uncontrolled
emission source
Pumps
Light 1 iquid service
Valves
Gas service
Light liquid service
Safety/relief valves
Gas service
Compressors
Inspection
interval
Monthly
Monthly
Monthly
Monthly
Monthly
Uncontrolled
emission
factor,
kg/hr
0
0
0
0
0
.120
.021
.010
.160
.440
b
Ac
0.87
0.98
0.84
0.69
0.84
Correction
factors
B
0.
0.
0.
0.
0.
d
95
95
95
95
95
Ce
0.98
0.98
0.98
0.98
0.98
Df
0.92
0.99
0.94
0.97
0.97
Control
efficiency
(AxBxCxD)
0.
0.
0.
0.
0.
75
90
74
62
76
Controlled9
emission
factor,
kg/hr
0.030
0.002
0.003
0.061
0.108
From Table 6-2.
bFrom Table 3-1.
Theoretical maximum control efficiency.6
Leak occurrence and reoccurrence correction factor - assum d to be 0.80 for yearly inspection, 0.90
for quarterly inspection, and 0.95 for monthly inspection.^
g
Non-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
average repair time yields a 0.98 yearly correction factor [365 - (15/2)] v 365.8
Imperfect repair correction factor - calculated as 1 - (f + F). Where f = average emission rate for
sources at 1000 ppm and F = average rate for emission sources greater than 10,000 ppm.9»10
^Controlled emission factor = uncontrolled emission factor x [1 - (A x B x C x D)].
-------
. TABLE 7-3. EMISSION FACTORS FOR SOURCES CONTROLLED UNDER REGULATORY ALTERNATIVE IV
en
Uncontrolled
emission source
Pumps
Inspection3
interval
Light liquid service None
Valves
Gas service
Monthly
Light liquid service Monthly
Safety/relief valves
Gas service
Compressors
Sampling connections
aFrom Table 6-2.
bFrom Table 3-1 .
°Theoretical maximum
Leak occurrence and
nuartprlu i n cnprt i nr
None
None
None
control effici
Uncontrolled
emission
factor,
kg/hr
0.120
0.021
0.010
0.160
0.440
0.015
ency .^
recurrence correction factor
i_ anH 0 QR for month! v in<;npr
Correction
factors
Ac
i_
NAh
0.98
0.84
NA-
NA
NA
- as sum
•tinn t^-
j
Bd
NA
0.95
0.95
NA
NA
NA
ed to
Q
ce
NA
0.98
0.98
NA
NA
NA
be 0.
£
D
NA
0.99
0.94
NA
NA
NA
80 for
Controlled9
Control emission
efficiency factor,
(AxBxCxD) kg/hr
O.O1
0.90 0.002
0.74 0.003
0.0
O.O1
0.0
yearly inspection, 0.90 for
eNon-instantaneous repair correction factor - for a 15-day maximum allowable repair time, the 7.5-day
average repair time yields a 0.98 yearly correction factor [365 - (15/2)] v 365.
Imperfect repair correction factor - calculated as 1 - (f -e- F). Where f = average emission rate for
sources at 1000 ppm and F = average rate for emission sources greater than 10,000 ppm. '
^Controlled emission factor = uncontrolled emission factor x[l -(AxBxCx D)].
Since the equipment associated with this regulatory alternative essentially eliminates fugitive
emissions, these correction factors are not applicable.
Emissions from pumps and conpressors equipped with double seals and vents to a 95 percent
control device are very small and are assumed to be zero for calculation purposes.
-------
TABLE 7-4. EXAMPLE CALCULATION OF VOC FUGITIVE EMISSIONS FROM MODEL
UNIT A UNDER REGULATORY ALTERNATIVE II
Number of
sources in
model unit9
(N)
Emission Source:0
Pumps
Light liquidd single
mechanical §eal
Light 1 iquid'-'dual
mechanical seal
Heavy liquid6 single
mechanical seal
Heavy liquid6 packed seal
In-line valves
Vapor service
Light liquidd service
Heavy liquid6 service
Safety/relief valves
Vapor service
Light liquid0" service
Heavy liquid6 service
Open-ended valves
Vapor service
Light liquid^ service
Heavy liquid6 service
Compressors
Sampling connections
Flanges
5
3
5
2
90
84
84~
11
1
1
9
47
48
1
26
600
Total
Emission
factor, b
kg/hr-source
(E)
0.044
0.044
0.020
0.020
0.003
0.004
0.0003
0.067
0.006
0.009
0.003
0.004
0.003
0.126
0.015
0.0003
emissions
Emissions
from sources.
kg/hr
(N x E)
0.220
0.132
0.100
0.040
0.270
0.336
0.025
0.737
0.006
0.009
0.027
0.188
0.014
0.126
0.390
0.180
2.800
Model units are characterized in Table 6-1.
Emission factors from Tables 3-1 and 7-1.
cSources in VOC service.
Light liquid service means that the fugitive emission source contains a
liquid which has a vapor pressure equal to or greater than 0.3 kPa at
20°C.
eHeavy liquid service means that the fugitive emission source contains a
liquid which has a vapor pressure less than 0.3 kPa at 20°C.
Open-ended valve factor is equivalent to the in-line valve factor because
capping the open end is assumed ,to eliminate emissions from this source.
7-6
-------
TABLE 7-5. ESTIMATED EMISSIONS AND EMISSION REDUCTIONS ON A MODEL UNIT BASIS'
Estimated emissi
(Mg/yr)
Regulatory Model unit
Alternative A B
I 67 260
II 24 94
III 21 80
IV 8 34
ons,b>c
C
800
290
250
106
Average percent
reduction from emissions
estimated under
Regulatory Alternative I
--
63
69
87
Average incremental
percent, reduction
in emissions
—
63
6
18
The emissions and percentage reductions presented in this table were calculated using the following:
• controlled and uncontrolled emission factors (see Tables 7-1, 7-2, and 7-3), and
• emission sources given in Table 6-1.
3A year is assumed to be equivalent to 8,760 hours.
C1.0 Mg/yr = 2200 pounds/yr
-------
3) annual replacement of the industry population based on a
twenty-year equipment life; and
4) annual growth rate of 5.9 percent for the industry.
Using these bases and the techniques presented in Appendix E, the total
number of model units in operation in 1981 were estimated to be 148.
18
In 1985 the total number of model units were estimated to be 831.
Under Regulatory Alternative I, total VOC fugitive emissions from
model units were estimated to increase from 35 to 199 gigagrams per
year (Gg/yr) during the same five-year (1981-1985) period (see Table 7-6).
In the same time period, implementation of Regulatory Alternative II
could be expected to reduce the baseline case (Regulatory Alternative I)
fugitive emissions by 63 percent. Implementation of Regulatory Alternative III
would reduce the baseline emissions by 69 percent. As Table 7-5 indicates,
Regulatory Alternative IV, the most stringent of all the alternatives,
would reduce the baseline emissions by about 87 percent.
7.2 IMPACT ON WATER QUALITY
In the absence of standards to reduce fugitive emissions of VOC from
SOCMI and under normal equipment operation, liquid leaks from various
equipment components could increase the quantity of wastewater generated
by a "typical" SOCMI facility. Under Regulatory Alternative I, liquid leaks
could originate from pumps and process valves in light or heavy liquid
service as well as valves on open-ended lines in light or heavy liquid service
and enter the wastewater system as runoff. Although the uncontrolled emission
rates for these sources are given in Chapter 3, the gas-liquid split of
these emissions is not defined. Consequently, the increase in wastewater
from SOCMI due to liquid leaks from potential fugitive emission sources
cannot be quantified.
Implementation of Regulatory Alternative II could reduce the wastewater
from a "typical" SOCMI facility by reducing the fugitive liquid emissions
resulting under Alternative I. The 'reduced emissions would be due to the
use of caps, plugs or second valves on open-ended lines in gas and light or
heavy liquid service. For example, caps, plugs, or second valves required
y-H
-------
under Alternative II would reduce the VOC fugitive emission rate from open-
ended lines in light or heavy liquid service from 0.01 kg/hr under Alternative
I to 0.004 kg/hr. This reduction would reflect a reduction in gaseous
emissions and liquid leaks. Since the gas-liquid split of the emission from
a given source is site specific, the impact of Alternative II on waste-
water from SOCMI cannot be quantified. However, it is likely that this
impact would be minor.
Implementation of Alternative III would result in impacts on wastewater
from SOCMI similar to those resulting from Alternative II. However, the
impacts under Alternative III would be more pronounced due to the more
frequent inspection intervals required by this alternative. The more
frequent intervals would reduce the VOC fugitive emission rate from valves
in light or heavy liquid service from 0.004 kg/hr under Alternative II to
0.003 kg/yr under Alternative III. Similarly, the fugitive emission rate
from pumps in light liquid service^would be 0.044 kg/hr under Alternative II
and 0.03 kg/hr under Alternative III. Consequently, the potential for the
production of liquid leaks which would be added to the wastewater from SOCMI
by possible fugitive emission sources would be less under Alternative III
than under Alternative II.
Of the alternatives being considered, Regulatory Alternative IV
could have the greatest impact on the quality of water that is discharged
from a "typical" SOCMI facility. Implementation of this alternative could
have positive (and possibly some negative) impacts on wastewater depending on
the specific control device requirements at each unit. Implementation of
Regulatory Alternative IV could reduce the amount of wastewater from a
SOCMI facility by reducing the fugitive liquid emissions resulting under
Alternative I. The reduction of these emission levels is primarily due to
the reduction of leaks from equipment in light liquid service, e.g., from
the use of double mechanical seals for pumps and closed loop sampling. Under
Regulatory Alternative IV, a double mechanical seal-degassing vent arrangement
reduces the emission rate of a pump seal in light liquid service under
Regulatory Alternative I from 0.12 to 0.0 kg/hr. A portion of this emission
reduction would be a reduction in liquids leaked to the ground or ditch.
However, the amount of liquids leaked to the ground or ditch that could enter
a plant wastewater system is not known.
7-9
-------
TABLE 7-6. TOTAL VOC FUGITIVE EMISSIONS FROM AFFECTED MODEL UNITS
FOR REGULATORY ALTERNATIVES
Total fugitive emissions estimated
Number of affected under Regulatory Al ternativ£>c
Year
1981
1982
1983
1984
1985
model units0 I
A B C (Gg/yr)
77 49 22 35.4
158 . 100 46 73.1
244 155 71 113.0
335 213 '=••• 97 155
432 274 125 199
II
(Gg/yr)
12.9
26.7
41.2
56.5
72.8
III
(Gg/yr)
11.0
22.8
35.2
48.3
62.1
IV
(Gg/yr)
4.6
9.5
14.8
20.2
26.0
The bases for estimating the number of model units,-as detailed in Appendix E, are:
an industry growth rate of 5.9 percent per year,
unit replacement based on a 20-year equipment life, and
• a base year (1980) total of 872 Model A, 554 Model B, and 252 Model C Units.
Estimated total VOC fugitive emissions from Model Units A, B, and C.
cDoes not include emissions from units in existence prior to 1981.
-------
Implementation of Regulatory Alternative IV could also result in a
negative impact on water quality due to the operation of a control device
which "captures" the fugitive VOC's. If a carbon adsorption device
were used to capture any VOC released at the degassing vent and if the
carbon is regenerated at the unit, a wastewater containing suspended solids
and some dissolved organics could be produced during the carbon regeneration
process. The use of a refrigeration process as the ultimate control device
could possibly result in a condensate containing dissolved organics. The
wastewater flow rates would be quite small and would generally be suitable
for treatment in the existing unit wastewater treatment process. Overall,
the impacts, both positive and negative, of Alternative IV on wastewaters
from SOCMI would be minnr.
7.3 IMPACT ON SOLID WASTE
In the absence of standards to reduce fugitive emissions of VOC from
SOCMI and under normal operation, solid wastes that could result from SOCMI
include replaced seals, packing, rupture disks, equipment components such
as pumps and valves, spent catalysts, and polymerization products. Metal
solid wastes such as mechanical seals, rupture disks and valve parts could
be sold as scrap metal to companies which can recycle the metal. This would
help to minimize the impact on solid waste. The quantity of used valve
packings and used batteries for monitoring instruments would not signifi-
cantly contribute to solid waste.
Implementation of Alternatives II and III would require the use of caps,
plugs, or second valves on open-ended lines in light or heavy liquid service,
and more frequent monitoring intervals. Implementing either of these
alternatives would have no greater impact on solid waste than Alternative I.
This is due to the relatively long life of caps, plugs, and second valves on
open-ended lines as well as the ability to sell discarded components such
as valves, mechanical seals, and rupture disks as scrap metal.
Implementation of Regulatory Alternative IV could result in the
generation of solid waste if carbon adsorption were used as a control
device and if the carbon were discarded instead of being regenerated.
However, the VOC emissions from the pump and compressor vents are small
streams, so that carbon requirements would be very low. Furthermore,
the carbon could be sent back to the manufacturer for regeneration,
thereby reducing the solid waste problem at the facility. It is antici-
pated that the manufacturer could incinerate or commercially dispose of
7-11
-------
any carbon that could not be regenerated (such as carbon fines) without
any serious environmental problems. Consequently, the negative impact of
implementing Alternative IV would be minor.
7.4 ENERGY IMPACT
Regulatory Alternatives II, III and IV call for passive controls on
equipment handling VOC streams (i.e., pump seals, process vent enclosures,
degassing vents, etc.); so implementing any of these alternatives will not
significantly increase the energy usa'je of a typical SOCMI plant. If a
control device such as a carbon adsorption system were used, steam (or another
hot regenerating medium) would be needed to regenerate the carbon at the unit;
however, the energy requirements would be quite small. The energy require-
ments of vapor recovery systems and of closed loop sampling would also be
small. Any of the alternatives would increase efficiency of raw material
usage. Because the raw materials for SOCMI are also energy sources, imple-
mentation of any of the alternatives being considered will result in a
positive energy impact.
The average energy value of the fugitive VOC emissions from SOCMI is
estimated to be approximately 31 x 10° joule/kg.'° The energy savings
resulting from the fugitive VOC emission reductions associated with
Alternatives II, III, and IV are presented in Table 7-7. Because Alterna-
tive IV is the most stringent, it will result in the greatest emission
reduction. As Table 7-7 indicates, implementation of this regulatory alter-
native would reduce the uncontrolled fugitive emissions by 173 Gg in the
fifth year and by a total of 520 Gg over a five-year period after implemen-
tation. These "recovered" VOC emissions have a total energy value of
no C
1.55 x 10I0 joules based on an average heating value of 31 x 10 joule/kg.
c- on
Assuming an energy value of 5.8 x 10° Btu per barrel of crude oil, the
energy value of the total fugitive emissions recovered over the five-year
period is approximately equal to 2.5 million barrels of crude oil under
Regulatory Alternative IV. This corresponds to an average daily savings of
1390 bbl/day of crude oil over the five-year period.
7-12
-------
TABLE 7-7. ENERGY IMPACT OF EMISSION REDUCTIONS FOR REGULATORY ALTERNATIVES
Year
1981
1982
1983
1984
1985
5-year
total
Reduction from baseline Energy value of emission
emissions under reductions under Crude oil equivalent
Regulatory Alternatives, Regulatory Alternatives, of emission reductions,
Gga terajoule13 thousand barrels
II III IV II III IV IIC IIIC IVC
22.4 24.4 30.8 694 756 955 113 124 156
46.4 50.3 63.6 1,440 1,560 1,970 235 255 322
71.8 77.8 98.2 2,230 2,410 3,040 364 394 497
98.3 106 135 3,050 3,290 4,180 498 538 683
127 137 173 3,940 4,250 5,360 644 695 876
366 396 500 11,350 12,270 15,500 1,855 2,005 2,530
Estimated total VOC fugitive emission reduction from Model Units A, B, and C.
Based on 1.55 x 1013 joules/kg21: This may be slightly over estimated if safety/
relief valves are controlled by a closed vent and flare system.
cBased on 5.8 x 106 Btu/bbl crude oil.
-------
7.5 OTHER ENVIRONMENTAL CONCERNS
7.5.1 Irreversible and Irretrievable Commitment of Resources
Implementation of any of the various alternatives is not expected
to result in any irreversible or irretrievable commitment of resources.
As previously noted, the regulatory alternatives should help to save
resources due to the energy savings associated with the reductions in
emissions.
7.5.2 Environmental Impact of Delayed Standards
As it was indicated above, implementation of the standards will
only have minor impacts on water and solid wastes. Consequently, delaying
the standards would have essentially no impact on these problems.
However, a delay in implementing the alternatives would have a greater
impact on air pollution and associated energy losses. The air and
energy impacts of delayed standards are shown in Table 7-7. The emission
reductions and associated energy savings shown would be irretrievably
lost at the rates shown for each of the five years.
7.6 REFERENCES
1. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979.
2. Rosebrook, D. D. (Radian Corporation.) Proceedings: Symposium on
Atmospheric Emissions from Petroleum Refineries. (Prepared for
U. S. Environmental Protection Agency.) Research Triangle Park,
N. C. Publication No. EPA-600/ 9-80-013. March 1980.
3. Reference 2.
4. Reference 1.
5. Reference 2.
6. Reference 1.
7. Reference 2.
8. Reference 2.
7-14
-------
9. Reference 1.
10. Reference 2.
11. Reference 1.
12. Reference 2.
13. Reference 2.
14. Reference 1.
15. Reference 2.
16. Memo from Muela, C. A., Radian Corporation, to Hudstvedt, K. C.,
EPArCPB. May 11, 1979. 1 p. Replacement rate of process unit in
the organic chemical industry.
17. McGraw-Hill Economics Department. The American Economy: Prospects
for Growth through 1991. New York, McGraw-Hill Publishing Company, 1976.
18. Letter from Smith, V. H., Research Triangle Institute, to Honerkamp, R.,
Radian Corporation. November 30, 1979. 1 p. Information about
baseline projections.
19. Memo from Blacksmith, J. R., Radian Corporation, to Hustvedt, K. C.,
EPA:CPB. December 15, 1979. 3 p. Information about energy value
of recovered product.
20. Petroleum Facts and Figures, 1971 Edition. Washington, D. C.,
American Petroleum Institute, 1971.
21. Reference 19.
7-15
-------
-------
8. COST ANALYSIS
8.1 COST ANALYSIS OF REGULATORY ALTERNATIVES
8.1.1 Introduction
The costs of implementing the regulatory alternatives for controlling
fugitive emissions of volatile organic compound (VOC) from the synthetic
organic chemicals manufacturing industry (SOCMI) are presented in the
following sections. Detailed descriptions of the model units and regulatory
alternatives treated in this cost analysis are presented in Chapter 6.
8.1.2 New Facilities
8.1.1.1 Capital Costs. The bases for the capital costs for the
model units are presented in Table 8-1. The capital cost estimates for each
model unit under each regulatory alternative are given in Table 8-2.
Regulatory Alternative I requires no control of VOC emissions. Consequently
there are no capital costs associated with this alternative.
The capital costs for the model units are the same under Regulatory
Alternatives II and III, since the only change is the monitoring frequency.
These costs include the purchase of two VOC monitoring instruments and caps
for all open-ended lines. It is assumed that one monitoring instrument
is used as a standby spare.
Under Regulatory Alternative IV, like II and III, two monitoring
instruments and caps for all open-ended lines would be purchased. In
addition, several other capital costs would be incurred. All single seal
pumps in light liquid service would require double mechanical seals at a
cost of $575/pump. A barrier fluid system ($1500/pump) would also be
required in conjunction with the douMe mechanical seals. Existing pumps
with double mechanical seals are assumed to have a barrier fluid system
already incorporated. Hence, there would be no additional capital
expenditure for the double seals or barrier fluid [system"."* "r-" V^tff
8-1
-------
TABLE 8-1. CAPITAL COST DATA
Item
Cost Value Used in Analysis
(last quarter 1978J)
Monitoring Instrument
Caps for open-ended lines
2 x 1250 = SBOO/model unit
45/line
Cost
Reference
line ins Irumpnt usnri as a spare 1,2
Rased on installation of a P.S 3,1,F>,ii,7
cm. screwed valve.3 Cost (1967) •
$12. Cost index = 2/H.I/l IJ.
Installation = 1 hour at SIS/hoiir.
Dual mechanical seals
575/pump (new)
Seal cost = $560. Sinqle seal
credit = $225. Shop
installation = $240.
350/pump (retrofit)
Seal cost = $560. Field
installation = t290.
Birrier fluid system for
dual mechanical seals
1500/pump
Pressurized reservoir system •
$700. System cooler - $800.
Pumps that have dual mechanical
seals without regulatory require-
ment may not have the co;t of a
barrier fluid system added. The
barrier fluid system is assumed to
be an Integral part of the seal
system.
10
Closed vrnts for dpqasslni)
reservoirs of corcprnssors
and dual se.i1 pumps
0530/coiiiprossor
Rased on Installation of a 12? in
Ipnfjt.h of 5.1 cm. '1 i.nnpt.pr,
srhedijlo 1f] cflrhon r.t.nts For compressors. The
cost for purups is basod on the
•TSSuiipUon that two pumps (such
as a pump aid its spare) are
ronnect.eri to a sinqlo floqassing vent.
aL1nes larger than J.5 em nay be flanged.
U estimated to o« !JO/Hne.
Installed cost For hl-nd flanges
Reproduced from
best available copy.
8-2
-------
TABLE 8-1 (cont.). CAPITAL COST DATA
Item
Cost Value Used in Analy-1s
(last quarter 1970$)
Cost Basis
Reference
Rupture disks for relinf
valves
1730,/rel ief valve (new)
f.ist (if ru|)turi> di'.k nsseiuMy: 13,11,1^,
one /.fp rm. mphire disk slainl"'.', 16,17,18,
slcel - $196; one 7.6 cm. rupture ' 19
disk holder, carbon steel = S32r>;
one 0.6 cm. pressure qauc installed up-
stream of the rupture disk. Cost
(1967) for one 7.6 cm. qate valve =
S'MO. Cost index = 27R. 1/113.
Installation = 10 hours at S15/hour.
To prevent damage to the relief valve
by disk fragments, an offset mounting
is required. Cost (1967) for one 10.2 cm.
Ice and one 10.2 rm. elhow - J7.30.
Cnst index - ?7P.1/113.
Installation = 8 hours at
3110/relief valve (retrnllt)
Costs for the rupture disk, holder,
and hlorV valve are the same as for
the new applications. An additional
cost is added to replace the derated
relief valve. No crpdil Ir, assuinp.1
fur the used relief valve. Cost for
one 7.6 cm pressure relief valve,
stainless steel hody and trim =
$500. Cost index = 2TR.1/113.
Installation = 10 hours at
20,21,22,21
Closed loop sampling
connections
460/samplinq connection
Ravd on installation of a f> >i
of P.S '-in. diiimelT, scl-i'dul
-------
TABLE 8-2. CAPITAL COST ESTIMATES FOR NEW MODEL UNITS
(thousands of.last quarter 1978 dollars)
Capital cost item I
Model Unit A
1. Monitoring instrument
2. Caos for open-ended 1 tries
3. Dual mechanical seals .
• Seals
• Installation c
4. Barrier fluid system for dual mech. seals
5. Vents for compressor degassing reservoirs
6. Vents for pump degassing r.ejervoirs
7. Rupture disks for relief valves
• Disks
• Holders, block valves, installation
8. Closed loop sampling connections
Total 0.0
Model Unit B
1. Monitoring instrument
2. Caps for open-ended lines
3. Dual mechanical seals >
• Seals
• Installation
4. Barrier fluid system for dual mech, sealsc
5. Vents for compressor degassing reservoirs
6. Vents for pump degassing reservoirs
7. Rupture disks for relief valves
• Disks
• Holders, block valves, installation
0. Closnd loop sampling connections
Tol.il n.o
Huilo 1 lln i I C
1. Monitoring instrument
2. Caps for open-ended 1 ines
3. Dual mechanical seals ,
• Seals
• Installation
4- Barrier fluid system for dual mech. sealsc
5. Vents for compressor degassing reservoirs
6. Vents for pump degassing reservoirs
7. Rupture disks for relief valves
• Disks
• Holders, block valves, installation
8. Closed loop sampling connections
Total ' 0.0
Regulatory alternative
II lit IV
8.50 8.50 8.50
4.68 4.68 4.58
1.68
1.20
7.5
6.53
26.1
2.14
16.8
12.0
13.2 13.2 87.1
8.50 8.50 8.50
18.7 18.7 18.7
6.36
4.56
28.5
13.1
94.7
8.19
64.4
47. B
'/'I.? ?.'! .2 ?
-------
Also, under Regulatory Alternative IV, compressor seals and pump
seals must have the seal oil degassing vents that are connected to a control
device such as a vapor recovery system or an enclosed combustion device. The
cost is estimated to be $6530 per compressor and $3265 per pump. This cost
is based on the assumption that one closed vent system is required for each
compressor. Since main pumps and spares are generally located in close
proximity to each other, one closed vent system would be required for each
pair of pumps. These costs are ba^sed on connecting the closed vent system
to an existing control device.
The costs of purchasing and installing rupture disks is $1590 per
relief valve. Rupture disks would be installed upstream of relief valves
in gas service. The cost includes the purchase of a shutoff valve to
allow the disk to be replaced after overpressure relief.
The closed loop sampling connection costs are based on an estimate of
$460 per sampling connection for installation of 6 meters of pipe and three
valves.
8.1.2.2 Annual Costs. With the implementation of Regulatory Alterna-
tives II, III, or IV, visual and/or instrument monitoring of potential
sources of fugitive VOC emissions will be required. A summary of the
requirements for the different alternatives is presented in Chapter 6.
Tables 8-3, 8-4, and 8-5 give the monitoring labor-hour requirements
for Regulatory Alternatives II, III and IV, respectively. The labor-hour
requirements were calculated by taking the product of the number of workers
needed to monitor a component (1 for visual, 2 for instrument), the time
required to monitor, the number of components in the model unit, and the
number of times the component is monitored per year. Monitoring labor costs
or Of. O~7
were then calculated based on $15 per hour. ' ' Regulatory Alternative III
would require the highest annual monitoring costs.
Leak repair labor is the cost of repairing those components in which
leaks develop after initial repair. Leaks may be discovered during the
8-5
-------
TABLE 8-3. ANNUAL MONITORING AND LEAK REPAIR LABOR REQUIREMENTS
FOR REGULATORY ALTERNATIVE II
Monitoring
Number of
components per
model unit
Source type ABC
Pumps (light liquid)
Single mechanical 5 19 60
seals
Dual mechanical 3 10 31
seals
Valves (in-line)
Gas 90 365 1117
Light liquid 84 335 1037
oo Safety/ ^el ief valves 11 42 13C
i (gas service)
o\
Valves on open-ended
1 inesh
Gas 9 37 115
Light liquid 47 189 581
Compressor seals 1 2 8
Type of*
monitoring
Instrument
Visual
Instrunent
Visual
Instrunent
Instrunent
Tr, = ; -.,--••' -i
Instrument
Instrument
Instrument
Monitoring
time,''
min
5
0.5
5
0.5
1
1
8
1
1
10
Times
monitored
per year
1
52
1
52
4
1
4
4
1
4
Leak repair
Estimated
Monitoring labor- number of . Repair Leak repair labor-
hours required0 leaks per year3 time, hours required6
A
1.0
2.2
1.0
1.3 •<
12.0
2.8
11.7
1.2
1.6
1 .3
B C A B C
3.2 10.0 1 1 3
8.2 26.0
1.7 5.2 112
I 4.3 13.4
49.0 149.0 4 15 45
11.2 34.6 3 9 25
".3 13:-. C
4.9 15.3 1 2 5
6.3 19.4 2 6 14
2.7 10.7 1 1 2
hrs ABC
80b 80 80 240
80b 80 80 160
1.13f 4.5 17.0 50.9
1.13f 3.4 10.2 28.3
O9 0 00'
1.13e 1.1 2.3 5.7
1.13e 2.3 6.8 15.8
40b 40 40 80
2 workers for instrument monitoring, 1 for visual. Ref. 20, p. 4-3.
bRef. 29.
^Monitoring labor-hours .= number of workers x number of components x time to monitor (total is minimum of 1 hr)
From Table 4-2.
Leak repair labor-hours = number of leaks x repair time.
*' "nd
°'17
It is assumed that these leaks are corrected by routine maintenance at no additional labor requirements. Ref. 3]
'«"<« "paired
-------
TABLE 8-4. ANNUAL MONITORING AND LEAK REPAIR LABOR REQUIREMENTS
FOR REGULATORY ALTERNATIVE III.
Monitoring
Number of
components per
model unit
Source type ABC
Pumps (light liquid)
Single mechanical 5 19 60
seals
Dual mechanical 3 10 31
seals
Valves (in-line)
Gas 90 355 1117
Light liquid 84 335 1037
Sa^etv'-e^er valves 11 42 130
(gas service)
CO
i
^ W:ve:- :r. c :-:-•• -ended
lines*
Gas 9 37 115
Light liquid 47 189 581
Compressor seals 1 2 8
Monitoring
Type of^ time, b
monitoring min
Instrument
Visual
Instrument
Visual
Instrument
Instrument
InstruneT*.
Instrument
Instrument
Instrument
5
0.5
5
0.5
1
1
a
1
1
10
Times
monitored
per year
12
52
12
52
12
12
12
12
12
12
Leak repair
Estimated
Monitoring labor- number of Repair Leak repair labor-
hours requ1redc leaks per yeard time, hours required6
A
10.0
2.2
6.0
1.3
36.0
33.6
35.2
3.6
18.8
4.0
B
38.0
8.2
20.0
4.3
146.0
134.0
134.4
14.8
75.6
8.0
C ABC
120.0 1 3 9
26.0
62.0 1 2 5
13.4
446.8 6 22 68
414.8 7 25 75
41C.O
46.0 1 3 7
232.4 4 14 42
32.0 1 1 2
nrs A B C
80 b 80 240 720
80 b 80 160 400
1.13f 6.8 24.9 76.8
1.13f 7.9 28.3 84.8
O9 0 0 0
1.13e 1.1 3.4 7.9
' 1.13e 4.5 15.8 47.5
40b 40 40 80
"<; workers for instrument monitoring, 1 for visual. Ref. 3?-
bRef. 13.
Monitoring labor-hours = number of workers x number of components x time to monitor (total is minimum of 1 hr^ .
From Table 4-2.
eLeak repair labor-hours = number of leaks x repair time.
r
'Weighted average based on 75 percent of the leaks repaired or-line, requiring 0.17 hour per repair, and on 25 percent of the leaks repaired
off-line, requiring 4 hours per repair. Ref. 34 .
^It is assumed that these leaks are corrected by routine maintenance at no additional labor requirements, fief. 3S.
The estimated number of leaks per year for open-ended valves is based on the same percent of sources used for in-rine valves. This represents
leaks occurring through the stem and gland of the open-ended valve. Leaks through the seat of the valve are eliminated hy addinq caps for
Renulatory Alternatives II, III, IV.
I Reproduced from
-------
TABLE 8-5. ANNUAL MONITORING AND LEAK REPAIR LABOR REQUIREMENTS FOR
riinUL,"TC''<,' ''LTE1' "THE IV.
Monitoring^
Number of
components per
model unit
Source type A
Pumps (light liquid)
Single mechanical 5
seals converted to
double seals
Dual mechanical ' 3
seals
Valves (in-line)
Gas 90
Light liquid 84
Safety/relief valves 11
(gss service)
CQ Valves on open-ended
C3 • "'
Gas 9
Light liquid 47
Compressor seals 1
B
19
10
jo5
335
42
37
189
2
C
60
31
1117
1037
130
115
581
8
Type ofa
monitoring
Instrument
Visual
Instrument
Visual
Instrument
Instrument
Instrument
—
Instrument
Instrument
Instrument
Monitoring
time, i>
min
5
0.5
5
0.5
1
1
8
1
1
10
Times
monitored
per year
Of
52
of
52
12
12
o-f
-
12
12
of
Monitoring labor-
hours requiredC
A
0
2.2
0
1.3
36.0
33.6
0
3.6
?8.8
0
B
0
8.2
0
4.3
146.0
134.0
0
14.8
75.6
0
C
0
26.0
0
13.4
446.8
414.8
0
46.0
232.4
0
Leak repair
Estimated
number of Repair Leak repair labor-
leaks per year^ time, hours required6
A B C hrs ABC
Of Of Of 80b 0 0 0
Of Of Of 80b 0 0 0
6 22 68 1.139 6.8 24.9 76.8
1 25 75 1.139 7.9 28.3 84.8
9f of of of>h ooo
1 3 7 1.139 1.1 3.4 7.9
4 14 42 I.139 4.5 15.8 47.5
ofo'ofiob o o o
2*orkersfor instrument nom'toring, I for 'visual. Re?. 36.
bRef. 37-
'Monitoring labor-hours = number of workers „ number of components x time to monitor (total
fro.T Table 4-2.
Leak repair labor-hours = number of leaks x repair time.
fHo monitoring or leak repair required because equipr^nt specifications eliminate leak potential
is a minimum o* 1 hrl
of the 1eaks repaired
.It is assumed that these leaks are corrected hy routine ™intenance at no additional labor requir
requirerents Ref 39
-------
periodic monitoring required by the regulatory alternatives. The number
of estimated leaks and the labor hours required for repair are given in
Tables 8-3, 8-4, and 8-5. Leak repair labor was calculated based on $15 per
hour.40'41'42 Maintenance labor costs would be greatest under Regulatory
Alternative III and least under Alternative IV. Costs would be reduced under
Alternative IV because the required installation of double mechanical
seals with seal oil degassing vents eliminates the most time-consuming
repair items.
Administrative and support costs were estimated at 40 percent of the
sum of monitoring and leak repair labor costs. Monitoring labor, leak
repair labor, and administrative/support costs are recurring annual costs
for each Regulatory Alternative.
8.1.2.3 Annual i zed Costs. The bases for the annual i zed control
costs are presented in Table 8-6. The annualized capital, maintenance,
and miscellaneous costs were calculated by taking the appropriate factor
from Table 8-6 and applying it to the corresponding- capital cost from
Table 8-2. The capital recovery factors were calculated using the
equation:
CRF .
(1 + i)n - 1
where i = interest rate, expressed as a decimal,
n = economic life of the component, years.
The interest rate used was 10 percent (last quarter 1978). The expected
life of the monitoring instrument wc:s 6 years compared to 10 years for other
control equipment components. Dual seals and rupture disks were assumed
to have a 2 year 1 ife.
The implementation of any of the Regulatory Alternatives (except I)
will result in the initial discovery of leaking components. It is
assumed that fewer leaks will be found at subsequent inspections. The
cost of repairing initial leaks was amortized over a 10-year period, since
this is a one-time cost. Repair of leaks found at subsequent inspections
was included as a recurring annual cost, in 8.1.2.2. The estimated
percentage of initial leaks per component is shown in Table 4-2. This
percentage was applied to the number of components in the model unit
under consideration. Fractions were rounded up to the next integer, since
in practice it is the whole valve, or seal, that is replaced and not just
part of one. The time required to repair each component type is given
8-9
-------
TABLE 8-6. DERIVATION OF ANNUALIZED LABOR, ADMINISTRATIVE,
MAINTENANCE AND CAPITAL CHARGES
1. Capital recovery factor for capital
charges
* Du*l s.eals and rupture disks
* Other control equipment
' Monitoring instruments
2. Annual maintenance charges
* Control equipment
• Monitoring instruments
3. Annual miscellaneous charges
(taxes, insurance, administration)
• Control equipment
• Monitoring instruments
4. Labor charges
5. Administrative and support costs to
implement regulatory alternative
6. Annualized charge for initial leak
repairs
0.58 x capital .
0.163 x capital1
0.23 x capital0
0.05 x capital
$2700e
0.04 x capitaU
0.04 x capital1"
$15/hour9
0.4 x (monitoring labor +
maintenance labor)!1
^(estimated number of leaking
components per model unit1 x
repair time1) x $15/hr9 x 1.4
x 0.1
Applies to cost of seals ($335 - incremental cost due to specification of
dual seals instead of single seals) and disk ($195) only. Two year life,
ten percent interest.
Ten year life, ten percent interes!-.. R-om Ref. 43, pp. IV-3,4.
cSix year life, ten percent interest. ; rom Ref. 44, pp. IV-9,10.
dFrom Ref. 45, pp. IV-3,4.
elncludes materials and labor for maintenance and calibration. Cost (last
quarter 1977) from Ref. 46, p. 4-3. Cost index = 221.7 * 209.1 (Ref. 47 and 48).
fFrom Ref. 49, pp. IV-3,4, 9, 10.
^Includes wages plus 40 percent for labor-related administrative and
overhead costs. Cost (last quarter 1977) from Ref. 50, pp. 4-4,5. Cost
index = 190.3 * 180.9 (Ref. 51 and 52).
hFrom Ref. 53, pp. IV-9,10.
in Table 8-7. ;
^Initial leak repair amortized for ten years at ten percent interest.
8-10
-------
in Table 8-7. The initial repair cost was determined by taking the product
of the number of initial leaks, the repair time, and the labor rate, $15
per hour. ' ' Forty percent wa'i added for administrative and support
costs. Finally, the total was multiplied by 0.163, the capital recovery
factor. As shown in Table 8-7, the cost of initial leak repair under
Regulatory Alternative IV is substantially less for each of the model
units than under Alternatives II and III. The main reason for this
reduction is the required installation of dual mechanical seals and
seal oil degassing vents that reduce the leak potential of pumps and
compressors. The repair time for a single pump or compressor seal is very
much greater than the repair time for a valve, so that a leak detection and
repair program for pumps and compressors would be more labor-intensive.
8.1.2.4 Recovery Credits. The annual VOC emissions, total emission
reductions, and annual recovered product credits for each model unit
under each Regulatory Alternative are shown in Table 8-8. Regulatory
Alternative I represents the uncontrolled emissions from each model unit.
The annual emission reduction was Calculated by subtracting the controlled
emission factor from the uncontrolled emission factor for each source.
To obtain an annual rate, the result was multiplied by 8760 hours per year.
The recovery credit was calculated at $360 per Mg of recovered product.
8.1.2.5 Net Annualized Costs. The net annualized costs, shown in
Tables 8-9, 8-10, and 8-11, were determined by subtracting the annual
recovered product credit from the tol;al cost before credit. For example,
Model Unit A, under Regulatory Alternative II has a net annualized credit
of $3300, as a result of $12,100 in costs and $15,400 for recovery
credits.
8.1.2.6 Cost Effectiveness. The cost effectiveness of each regula-
tory alternative for each model unit is shown in Table 8-12. Regulatory
Alternatives II and III have a net annualized credit for all model units,
and cost effectiveness numbers are negative. Since Regulatory Alternative
IV is the only one with a positive net cost, comparisons of cost
effectiveness in the normal sense are meaningless. The highest cost of
VOC control under Regulatory Alternative IV is for model unit A. Although
8-71
-------
TABLE 8-7. LABOR-HOUR REQUIREMENTS FOR INITIAL LEAK REPAIR
Regulatory alternative II Regulatory alternative III
Number of
components
per model
unit
Source type
Pumps (1 ight liquid)
Single mechanical seal
Dual mechanical seals]
Valves (in-line)
Gas
Light liquid
Safety/relief valves3
(gas service)
CO Valves on open-ended lines'
i
— ' Gas
ro
Light liquid
Compressor seals
A
5
3
90
84
11
q
47
1
B C
19 60
10 31
365 1117
335 1037
42 130
37 115
189 581
2 8
Estimated
number of
initial
leaks b
ABC
2 5 14
1 3 8
9 37 112
11 41 125
000
1 4 12
6 23 70
1 1 3
Repair
time ,
hrs
80C
80C
1.13d
1.13d
0
1.13d
1.13d
40C
Estimated
Labor-hours ""niti^
ABC ABC
160 400 1120 2 5 14
80 240 640 138
10 42 127 9 37 112
12 46 141 11 41 125
00 0000
1 5 14 1 4 12
7 26 79 6 23 70
40 40 120 1 1 3
Repair
time,
hrs
80C
80C
1.13d
1.13d
0
1.13d
1.13d
40C
Regulatory alternative IV
Estimated
number of
Labor-hours initial Repair
required leaks time,
ABC A
160 400 1120 Oe
80 240 640 Oe
10 42 127 9
12 46 141 11
0 0 0 Oe
1 5 14 1
7 25 79 6
40 40 120 Oe
B
oe
oe
37
41
Oe
4
23
Oe
C hrs
Oe 80 C
O6 80C
112 1.13d
125 1.13d
Oe 0
12 1.13d
70 1.13d
Oe 40
Labor- hours
required
ABC
000
COO
10 42 127
12 46 141
000
1 5 14
7 26 79
000
It is assumed that these leaks are corrected by routine maintenance at no additional labor requirements. Ref. 58.
Based on tie percent of sources leaking at > 10,000 ppm. From Table 4-2.
cRef. 59.
Weighted average based on 75 percent of the leaks repaired on-line, requiring 0.17 hours per repair, and on 25 percent of the leaks repaired off-line,
requiring 4 hours per repair. Ref. 60.
eNo maintenance required because equipment specification eliminates leak potential.
The estimated number of initial leaks for open-ended valves is based on the same percentage of sources used for in-line valves. This represents leaks
occurring through the stem and gland of the open-ended valve. Leaks through the valve seat are eliminated by adding caps for Regulatory Al terr.atives
II, III, IV.
-------
TABLE 8-8. RECOVERY CREDITS.
CO
Regulatory
alternative
I
II
I ,
voc
emissions,
Mq/yr
67.2
24.5
20.8
8.46
Model unit A
Emission reduction
from uncontrolled,
Mq/yr
--
42.7
46.4
53.7
Recovered^
product
value,
$/yr
--
15,400
21,100
VOC
emissions,
Mq/yr
257
93.7
79.8
34.3
Model unit B
Emission reduction
from uncontrolled,
Mq/yr
--
163
177
223
Recovered3
product
value,
S/yr
--
58,800
63, so:
80, 2C:
voc
emissions,
Mq/yr
800
293
249
106
Model unit C
Emission reduction
from uncontrolled,
Mq/yr
--
507
551
694
Recovered*
product
value,
$/yr
--
182.00C
250,001
Last quarter 1978 dollars. Based on an average price of $360/Mg. Ref. 61.
-------
TABLE 8-9. ANNUALIZED CONTROL COST ESTIMATES FOR MODEL UNIT A
(thousands of last quarter 1978 dollars).
Cost item
Annualized capital charges
1. Control equipment
a. Instrument
b. Caps
c. Dual seals
• Seals
• Installation t,
d, Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating costs
1. Maintenance charges
a. Instrument
b. Caps •
c. Dual seals
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks •
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Dual seals
d. Barrier fluid system
(; . Vents - Blimps and compressor'.
ft. Rupture disks
g. Closed toop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative and support
Total before credit
Recovery credits
Net annual 1 zed cost8
Regulatory alternative
•I II III
1.96 1.96
.763 .763 '
1 .06 . 1.06
2.70 2.70
.234 .234
.340 .340
.111/ .187
0.54 2.26
3.17 3.30
1.51 2.22
0.0 12.1 15.0
0.0 15.4 16.7
0.0 (-3.3) (-1.7)
IV
1.96
.763
0.974
.196
1.22
5.32
1.24
2.74
1.96
0.10
2.70
.234
0.144
.375
1.63
.950
0.60
.340
.187
.115
0.30
1.31
.75U
.48
1.43
0.304
0.692
29.0
21.1
7.9
aCost 1s for back-to-back arrangement.
bPressur1zed system.
cBased on 40 percent of monitoring plus leak repair labor. Ref. 62.
Based on an average price of $360/Mg. Ref. 63.
e(-xx)=-> net credit
8-14
-------
TABLE 8-10. ANNUALIZED CONTROL COST ESTIMATES FOR MODEL UNIT B
(thousands of last quarter 1978 dollars)
Cost item
Annualized capital charges
1. Control equipment
a. Instrument
b. Caps
c. Dual seals
• Seals
• Installation .
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating costs
1. Maintenance charges
a. Instrument
b. Caps
c. Dual seals
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed Idop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Dual seals
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c. Administrative ana support
Total before credit
Recovery credits
Net annual 1zed cost*
Requlatory
I II
1.96
3.05
2.73
2.7
.935
0.31
.7'1B
2.04
3.54
2.23
O.I) 20..)
0.0 '.ili.ii
0.0 (-38.5)
alternative
III
1.96
3.05
2.73
2.7
.935
0.34
.718
8.75
7.69
6.58
:i !).'">
n3.lt
(-ZU.3)
IV
1.96
3.05
3.69
.743
4.65
17.6
4.75
10.5
7.79
0.41
2.7
.935
.546
1.42
5.39
3.63
2.39
0.34
.74a
.437
1.14
4.31
2.90
1.91
5.74
1.09
2.73
93,'j
IIU.2
13.3
aCost 1s for back-to-back arrangement.
Pressurized system.
cSee footnote from proceeding Table 8-9, Ref. 64.
d3ased on an average price of $360/Mg. Ref. 65
e(-xx)—^. net credit
8-15
-------
TABLE 8-11. ANNUALIZED CONTROL COST ESTIMATES FOR MODEL UNIT C
(thousands of last quarter 1978 dollars)
Cost item
Annual! zed capital charges
1 . Control equipment
a. Instrument
b. Caps
c. Dual seals
• Seals
• Installation b
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
g. Closed loop sampling
2. Initial leak repair
Operating Costs
1 . Maintenance charges
a. Instrument
b. Caps
c. Dual seals
d. Barrier fluid system
e. Vents - pumps and compressors
f. Rupture disks
g. Closed loop sampling
?. Miscellaneous (taxes, insurance,
administration)
-------
TABLE 8-12. COST EFFECTIVENESS FOR MODEL UNITS
(last quarter 1978 dollars)
Model unit fta
Regulatory alternative
Total capital cost ($1000)
Total annualized cost ($1000)
Total annual recovery credit ($1000)
c» Net annualized cost ($1000)d
Total VOC reduction (Mg/yr)
Cost effectiveness
(annual $/Mg VOC)°
I II
0.0 13.2
0.0 12.1
0.0 15.4
0.0 (-3.3)
0.0 42.7
(-77.3)
III
13.2
15.0
16.7
(-1.7)
46.4
(-36.6)
IV
87.1
29.0
21.1
7.9
58.7
135.
Model unit Bb
I II
0.0 27.2
0.0 20.3
0.0 58.8
0.0 (-38.5)
0.0 163
- (-236.)
Ill
27.2
35.5
63.8
(-38.3)
177
(-160.)
IV
295
93.5
80.2
13.3
223
59.6
Model
I II
0.0 66.0
0.0 48.3
0.0 182.
0.0 (-134.)
0.0 507
- (-264.)
unit Cc
III
66.0
95.1
198.
(-103.)
551
(-187.)
IV
9]}
283
250.
33.0
694
47.6
a52 percent of the units in the SOCMI are similar to Model Unit A. Ref. 68.
b33 percent of the units in the SOCMI are similar to Model Unit B. Re*. 69.
C15 percent of the units in the SOCMI are similar to Model Unit C. Ref. 70.
. (-xx) = Control method net credit
-------
this cost ($135/Mg) is much larger than the cost for model unit C ($48/Mg),
the net annualized cost for model unit A is only $7900. This amount is
insignificant compared to the annual operating cost of the process unit
itself.
8.1.3 Modified/Reconstructed Facilities
8.1.3.1 Capital Costs. The bases for determining the capital costs
for modified/reconstructed facilities are presented in Table 8-1. The
capital costs for these units are the same under Regulatory Alternatives
II and III as are those for new units. There are no costs associated
with Alternative I. The capital costs for the monitoring instruments,
the caps for open-ended lines, the barrier fluid systems, the vents for
degassing reservoirs, and the closed loop sampling connections are also
the same as for new units.
The estimated cost of retrofitting dual mechanicals seals for
single seal pumps was estimated at $850 per pump. This figure includes
$560 for a new back-to-back dual mechanical seal plus $290 labor for field
installation.
Rupture disks for relief valves, required under Regulatory Alterna-
tive IV, were estimated to cost $2970 per relief valve. The original
relief valve must be replaced with a larger relief valve. The cost for a
new valve was included in the cost estimates. Credit for the removed valve
was not included.
The total capital cost estimates for modified/reconstructed facilities
are presented in Table 8-13. As noted above, the costs associated with
Regulatory Alternatives 1, II, and III are the same as for new units.
8.1.3.2 Annualized Costs. The annualized control costs for
modified/reconstructed units, presented in Table 8-14, are derived from
\"
the same basis as new units (see Table 8-2). The only changes from new I
unit costs occur under Regulatory Alternative IV because of the increased
capital costs for dual mechanical seals and rupture disks. The recovered I
product credits for the modified/reconstructed units are the same as for '
the new model units.
8-18
-------
TABLE 8-13. CAPITAL COST ESTIMATES FOR MODIFIED/
RECONSTRUCTED FACILITIES
(thousands of last quarter 1978 dollars)
1.
2.
3.
Capital cost item3
Monitoring instrument
Caps for open-ended lines
Dual mechanical seals0
• Seals
• Installation
Regulai
A
8.5
4.68
2.8
1.45
toj^y alternative
Model unit
B
8.5
18.7
10.6
5.51
IVb
C
8.5
57.5
33.6
17.4
4. Barrier fluid systems for double 7.50 28.5 90.0
mechanical seals
5. Vents for compressor degassing 6.53 13.1 52.2
reservoirs
6. Vents for pump degassing 26.1 94.7 297
reservoirs
7. Rupture disks for relief
valves
• Disks 2.14 8.19 25.4
• Holders, block valves, 16.8 64.4 199
installation
• Replacement relief valve and 15.2 58.0 179
installation
8. Closed loop sampling connections 12.0 47.8 147
Total 104 358 1107
aFrom Tables 6-1 and 8-1.
For Regulatory Alternatives I, II, III the capital costs for modified/
reconstructed facilities are the same as for new units (Table 8-2).
°Cost is for back-to-back arrangement.
Pressurized system.
8-19
-------
TABLE 8-14. ANNUALIZED CONTROL .COST ESTIMATES FOR MODIFIED/- - -
.^CONSTRUCTED'M.QDFL UNITS UNDER REGULATORY ALTERNATIVE IVa
(thousands of last quarter 1978 dollars)
Model . Model Model
Cost Item unit A unit Bc unite
Annuallzed capital charges
1. Control equipment
a. Instrument
b. Caps
c. Dual seals6
• Seals
• Installation ,
d. Barrier fluid system
e. Vents for pumps and compressors
f. Rupture disks
• Disks
• Holders, etc.
• Relief valves
g. Closed loop sampling
2. Initial leak repair
Operating costs
1. Maintenance charges
a. Instrument
b. Caps
c. Dual seals
d. Barrier fluid system
e. Vents for pumps and compressors
f. Rupture disks
g. Closed loop sampling
2. Miscellaneous (taxes, insurance,
administration)
a. Instrument
b. Caps
c. Dual seal:
d. Barrier fluid system
e. Vents for pumps and compressors
f. Rupture disks
g. Closed loop sampling
3. Labor
a. Monitoring labor
b. Leak repair labor
c- Administrative and support9
Total before credit
Recovery credits
Net annualized cost
Total VOC reduction (Mg/yr)
Cost effectiveness ($/Hg VOC)
1.96
.763
1.62
.236
1.22
5.32
1.24
2.74
2.48
1.96
0.10
2.70
.234
0.213
.375
1.63
1 .71
.60
.340
.187
0.170
0.30
1.31
1.37
.430
1.43
0.304
0.692
33.7
21.1
12.6
58.7
215.
1.96
3.05
6.15
0.898
4.65
17.6
4.75
10.5
9.45
7.79
0.41.
2.70
.935
0.806
1.42
5.39
6.53
2.39
.340
.748
.644
1.14
4.31
5.22
1.91
5.74
1 .09
2.73
111 .
80.2
30.8
223.
138.
1.96
9.37
19.5
2.84
14.7
56.9
14.7
32.4
29.2
24.0
1.23
2.70
2.88
2.55
4.50
17.5
20.2
7.35
.340
2.30
2.04
3.60
14.0
16.2
5.08
17.7
3.25
8.38
338.
250.
88.
694
127.
For Regulatory Alternatives I, II, III, the annualized control costs and cost
effectiveness for modified/reconstructed facilities are the same as for new units
(Tables 8-7, 8-8, 8-9).
52 percent of existing units are similar to Model Unit A. Ref. 71.
C33 percent of existing units are similar to Model U-iit B. Ref. 72.
15 percent of existing units are similar to Model Unit C. Ref. 73.
eCost
-------
8.1.3.3 Cost Effectiveness. The cost effectiveness figures for
modified/reconstructed facilities are also shown in Table 8-14. The cost
effectiveness under Regulatory Alternatives I, II, and III is the same as
for the new model units. The cost effectiveness under Regulatory Alter-
native IV is a net cost of $215 per Mg for model unit A, $138 for model
unit B, and $127 for model unit C.
8.1.4 Projected Cost Impacts
The regulatory alternatives are assumed to go into effect by 1981,
using 1980 as the base year. The industry is estimated to grow at a rate
of 5.9 percent/" SOCMI facilities are estimated to be replaced at a rate
based on a 20-year equipment life (see Appendix E). The estimated numbers
of projected new units are presented in Tables 7-5, 7-6, and 7-7. The
estimated costs to the industry for the years 1981 through 1985 are
presented in Tables 8-15 through 8-17. Capital costs shown are only for
units which begin operation in the indicated year. All other costs shown
are for all units subject to NSPS in the indicated year.
8.2 OTHER COST CONSIDERATIONS
Environmental, safety, and health statutes which are applicable to
SOCMI plants are listed in Table 8-18. The provisions, requirements, and
regulations listed are those which may cause an outlay of funds by an
organic chemical manufacturer.
Specific costs of each of these provisions or requirements to the
industry defined as SOCMI were unavailable. Total costs to SOCMI for
complying with environmental, safety and health standards were also
unavailable.
The entire chemical industry is planning to spend an estimated $639
million on pollution control in 1979 according to a McGraw-Hill Survey.77
Although this is a sizeable sum of money, the industry has enjoyed three
decades of rapid growth and high profits. The economic health of the
industry is better than that of many other industries.78 [he substantial
pollution problems encountered in the industry and the large expenditures
8-21
-------
ro
TABLE 8-15. NATIONWIDE COSTS FOR THE INDUSTRY UNDER REGULATORY ALTERNATIVE II
(last quarter 1978 dollars)
Cost item3
Total capital cost ($1000)b
Total annualized cost ($1000)c
Total annual recovery credit ($1000)
Net annualized cost ($1000)d (
1981
3,800
2,990
8,070
-5,080)
1982
4,040
6,160
16,700
(-10,500)
1983
4,280
- 9,530
25,800
(-16,300)
1984
4,490
13,100
35,300
(-22,200)
1985
4,790
16,800
45,500
(-28,700)
-aFrom Tables 3-2, 8-9, 8-10, 8-11.
Capital costs for model units which begin operation in the years shown.
Annualized costs for all model units subject to NSPS in the years shown.
(-xx) "* net credit
-------
TABLE 8-16. NATIONWIDE COSTS FOR THE INDUSTRY UNDER REGULATORY ALTERNATIVE III
(last quarter 1978 dollars)
'--'-
To til
• >
Total
Cost item3
capital cost ($1000)b
annualized cost ($1000)°
1981
3,800
4,990
1
4
10
982
,040
,300
1983
4
15
,280
,900
1984
4,490
21,800
1985
4
23
,790
,100
Total annual recovery credit ($1000) 8,770
18,100
28,000
38,400
49,400
00
CO
Net annualized cost ($1000)
(-3,780) (-7,800) (-12,100) (-16,600) (-21,300)
From Tables 7-6, 8-7, 8-8, 8-9.
Capital costs for model units which begin operation in the years shown.
"Annualized costs for all model units subject to NSPS in the years shown.
(-xx) =?> net credit
-------
ro
TABLE 8-17. NATIONWIDE COSTS FOR THE INDUSTRY UNDER REGULATORY ALTERNATIVE IV
(last quarter 1978 dollars)
Cost itema
Total capital cost ($1000)
Total annualized cost ($1000)°
Total annual recovery credit
($1000)
Net annualized cost ($1000)d
1981
41,200
13,000
11,100
1,900
1982
44,000
27,000
22,900
4,100
1983
46,500
41,700
35,400
6,300
1984
48,700
57,100
48,500
8,600
1985
51,900
73,500
62,400
11,100
aFrom Tables 7-6, 8-7, 8-8, 8-9.
Capital costs for model units which begin operation in the years shown.
°Annualized costs for all model units subject to NSPS in the years shown.
(-xx) =£> net credit
-------
TABLE 8-18. STATUTES THAT MAY BE APPLICABLE TO SOCMI
Statute
Applicable provision, regulation or
requirement of statute
Approximate cost incurred due to
enactment of statute
Model unit
Industry
Clean Air Act and Amendments
Clean Water Act (Federal
Water Pollution Act)
CO
i
PO
tn
Resource Conservation and
Recovery Act
State implementation plans Total
National emission standards for hazardous
air pollutants
Benzene fugitive emissions
New source performance standards
Air oxidation
Volatile organic liquid storage
PSD construction permits
Non-attainment construction permits
Discharge permits Total
Effluent limitations guidelines
New source performance standards
Control of oil spills and discharges
Pretreabnent requirements
Monitoring and reporting
Permitting of industrial projects that
impinge on wetlands or public waters
Environmental impact statements
Permits for treatment, storage, and Total
disposal of hazardous wastes
Establishes system to track hazardous
wastes
Establishes recordkeeping, reporting,
labelling and monitoring system for
hazardous wastes
Superfund
$249 million3
$414 million
$200 million0
Superfund-less than 22 of profits
or $200 mill ion maximum annual
rate on petrochemical
feedstock^
Production costs for the industry
are expected to increase by an
average of 0.6% and a maximum
of 5%.e
Toxic Substances Control
Act
Premanufacture notification
Labelling, recordkeeping
Reporting requirements
Toxicity testing
Total
S100-200 million per year
Preinventory notification cost:
$1200-1500 per chemical
(Continued)
-------
TABLE 8-18. (Cont.
Statute
Applicable provision, regulation or
requirement of statute
Approximate cost incurred due to
enactment of statute
Model unit
Industry
Occupational Safety and Health
Act
Walking-working surface standards
• Means of egress standards
• Occupational health and environmental
control standards
• Hazardous material standards
• Personal protective equipment standards
•General environmental control standards
• Medical and first aid standards
•Fire protection standards
• Compressed gas and compressed air
equipment
Welding, brazing, and cutting standards
Total
$220/year per worker
CO
8?
Coastal Zone Management Act
Power Plant and Industrial
Fuel Use Act
States may veto federal permits for
plants to be sited in coastal zone
Prohibits new, major, industrial power-
plants which utilize fuel oil or
natural gas
National Environmental Policy
Act
Requires environmental impact statements
Safe Drinking Water Act
Requires underground injection control
permits
Marine Sanctuary Act
Ocean dumping permits
Recordkeeping and reporting
Expenditure, by entire chemical industry, on air pollution control; SOCMI's portion of expenditure not delineated. (Ref. 80.!
Expenditure, by entire chemical industry, on water pollution control; SOCHI's portion of expenditure not delineated. (Ref. 81
cCost reflects entire organic industry; SOCMI's cost not delineated. (Ref. 82).
dCost reflects entire organic industry; SOCMI's cost not delineated. (Ref.83,84-).
eCost reflects entire organic industry; SOCHI's cost not delineated. (Ref. 85).
Cost incurred by entire chemical industry; SOCMI's portion, of expenditure not delineated. (Ref. R6).
9Cost incurred by entire chemical industry; SOCMI's portion of expenditure not delineated. (Ref. 87).
"Cost incurred by entire chemical industry; SOCMI's portion of cxpeivi-;turc not delineated. (Ref. 88).
-------
necessary for their solution are expected to affect the smaller firms more
adversely than the larger firms. However, few plant closings are expected
79
due solely to costs of compliance with standards and regulations.
The costs incurred by SOCMI in complying with all health, safety,
and environmental requirements are not expected to prevent compliance
with the proposed NSPS for fugitive emissions.
8-27
-------
8.3 REFERENCES
1. Letter and attachments from Amey, G. C., Century Systems Corporation,
to Serne, J., Pacific Environmental Services. October 17, 1979.
3 p. Cost data for VOC monitoring instrument.
2. Erikson, D. G. and V. Kalcevic. (Hydroscience, Inc.) Emissions
Control Options for the Synthetic Organic Chemicals Manufacturing
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-2577. February 1979.
p. IV-9.
3. Peters, M. S. and K. D. Timmerhaus. Plant Design and Economics for
Chemical Engineers, Second Edition. New York, McGraw-Hill Book
Company, 1968.
4. Kohn, P. M. CE Cost Indexes Maintain 13-Year Ascent. Chemical
Engineering. 85(11):189-190. May 8, 1978.
5. Economic Indicators. Chemical Engineering. 87(1):7. January 14,
1980.
6. U. S. Environmental Protection Agency. Control of Volatile Organic
Compound Leaks from Petroleum Refinery Equipment. Research Triangle
Park, N. C. Publication No. EPA-450/2-78-036. June 1978. p 4-5.
7. Economic Indicators. Chemical Engineering. 8£(2):7. January 15,
1979.
8. Reference 2, p. IV-3.
9. Reference 2, p. IV-3.
10. Reference 2, p. IV-3.
11. Reference 2, pp. IV-8,9.
12. Reference 2, pp. IV-8,9.
13. Reference 2, p. IV-8.
14. Reference 3, pp. 450-452.
15. Reference 4.
16. Reference 5.
17. Reference 7. .
8-28
-------
18. Reference 8.
19. Reference 3, p. 452.
20. Reference 8.
21. Reference 4.
22. Reference 5.
23. Reference 7.
24. Reference 2, p. IV-8.
25. Reference 5.
26. Reference 7.
27. Reference 8.
28. Reference 6, p. 4-3.
29. Letter and attachments from Johnson, J. M., Exxon Company, to
Walsh, R. T., EPA:CPB. July 28, 1977. 14 p. Review of "Control
of Hydrocarbon from Miscellaneous Refinery Sources" report.
30. Reference 2, p. B-12.
31. Reference 29.
32. Reference 6, p. 4-3.
33. Reference 29.
34. Reference 2, p. B-12.
35. Reference 29.
36. Reference 6, p. 4-3.
37. Reference 29.
38. Reference 2, p. B-12.
39. Reference 29.
40. Reference 5.r——-—. ._
•'••.i'r'-'.i •; •iii'J>^'.V^3''1'r''' ~J—"~~~
- 41.' Refer'en'ee •'•$>.'>•
8-29
-------
42. Reference 8.
43. Reference 2, pp. IV-3,4.
44. Reference 2, pp. IV-9,10.
45. Reference 2, pp. IV-3,4.
46. Reference 6, p. 4-2.
47. Reference 5.
48. Reference 8.
49. Reference 2, pp. IV-3,4,9,10.
50. Reference 6, pp. 4-4,5.
51. Reference 5.
52. Reference 8.
53. Reference 2, pp. IV-9,10.
54. Reference 5.
55. Reference 7.
56. Reference 8.
57. Letter from Smith, V. H., Research Triangle Institute, to Honnerkamp, R.,
Radian Corporation. November 30, 1979. 1 p. Information about
baseline projections.
58. Reference 29.
59. Reference 29.
60. Reference 2, p. B-12.
61. Reference 57.
<<;'. R(?|«M-3. Reference 57.
64. Reference 2, pp. IV-9,10.
65. Reference 57.
8-30
-------
66. Reference 2, pp. IV-9, 10.
67. Reference 57.
68. Reference 2, p. IV-1.
69. Reference 2, p. IV-1.
70. Reference 2, p. IV-1.
71. Reference 2, p. IV-1.
72. Reference 2, p. IV-1.
73. Reference 2, p. IV-1.
74. Reference 2, p. IV-1.
75. Reference 57.
76. Letter from Smith, V., Research Triangle Institute, to Honnerkamp, R.,
Radian Corporation. August 13, 1979. 1 p. Predicted growth rate
of the SOCMI industry.
77. News Flashes. Chemical Engineering. i36_(12):77. June 4, 1979.
78. Environmental Quality: The Seventh Annual Report of the Council on
Environmental Quality. Washington, D. C., U. S. Government Printing
Office. December 1976.
79. Environmental Quality: The Ninth Annual Report of the Council on
Environmental Quality. Washington, D. C., U. S. Government Printing
Office. December 1978.
80. Reference 78.
81. Reference 78.
82. U. S. Environmental Protection Agency. Solid Waste Facts, A Statistical
Handbook. Washington, D. C. Publication No. SW-694. August 1978.
16 p.
83. EPA Charges Chemical Trade Seeks "Lowest Denominator" as Its Position
on Superifund. Chemical Marketing Reporter. 216(10) :3. September 3,
1979.
84. Carter Accepts Tough Version of "Super Fund"; Would Cost Industry
$1.6 Bi 1,1; ion jfpr.,Cleanup., .Chemical. .Marketing. Reporter. 215(25) ;3.
June is,fg'fgf''•-'''^TO-' "<*•••• • ••: •
8-31
-------
85. Reference 78.
86. Reference 78.
87. Reproposal of Premanufacture Notice Form and Provisions of Rules.
Federal Register. 44(201):59764. October 16, 1979.
88. Arthur Anderson & Company. Cost of Government Regulation Study for
the Business Roundtable. Washington, D. C. March 1979. p. 8-6.
8-32
-------
9. ECONOMIC ANALYSIS
9.1 INDUSTRY PROFILE
9.1.1 Introduction
The synthetic organic chemicals manufacturing industry (SOCMI) has
been defined as the producers of synthetic organic chemicals, listed in
Appendix F. This profile gives a general qualitative description of the in-
dustry, supported by quantitative information wherever possible. Because
SOCMI does not directly correspond to industrial classifications used for re-
porting information by secondary data sources, a weighting technique was
used to develop industry statistics (see Appendix El).
Synthetic organic chemicals (SOCs) are substances containing at least
carbon and hydrogen. They exhibit three basic molecular structures: ali-
phatic or acyclic, cyclic, and combinations of aliphatic and cyclic. Acyclic
compounds are composed of groups of atoms arranged in a straight chain.
Examples are alcohols, ethers, ketones, and carbohydrates. Cyclic compounds
have the atoms of their component elements arranged in the form of a closed
ring. Examples include aromatic hydrocarbons, napthenes, and thiazoles.
Certain amino acids and terpene hydrocarbons represents combinations of
cyclic and aliphatic molecular structures.
SOCMI chemicals may be used as primary feedstocks, chemical intermedi-
ates, or end use chemicals. Primary feedstocks are produced from crude raw
materials and used in the manufacture of other chemicals. Chemical interme-
diates are the product of primary feedstocks and are also used to produce
other chemicals. End use chemicals are products of chemical intermediates
and/or primary feedstocks and are used either as final goods or as inputs to
production processes outside the chemical industry. Many synthetic organic
chemicals are used, .in .mo.re^than; one of these categories. Figure 3.1 illus-
'(--•j^l^Jjj.c' •.••,. i j • •>
tmtos the general relationships amonci the various organic chemical s.
9-1
-------
Detailed flow charts identifying inputs and product use for many of the SOCMI
o
chemicals have been presented elsewhere.
9.1.2 Production Processes and Capacities
Most of the SOCMI chemicaTs produced in the United States are derived
from crude petroleum and natural gas. Oil, shale, coal, and biomass (non-
prehistoric plant tissue) are also sources of primary feedstocks.3 A wide
variety of processes are used to manufacture the synthetic organic
chemicals included in the definition of SOCMI. Frequently individual
chemicals can be manufactured in several different ways. Consequently, as
relative prices chanqe, chemical producers may alter the mix of primary
feedstocks used to produce SOCs.
After chemical feedstocks are manufactured from petroleum, natural gas,
and other raw materials, they are processed into chemical intermediates and
end use chemicals. Some of the chemicals included in SOCMI are the product
of a simple distillation process, while others are produced from a series
of cracking processes.
In 1976 Organic Chemical Producers Data Base^ reports 1,270 units pro-
ducing SOCMI chemicals in the United States.* Table 9-1 presents a distri- f.
bution of those units and estimated capacity by state. New Jersey, Texas,
and California have the largest number of units producing SOCMI chemicals.
Texas and Louisiana have the largest total production capacities. These
states are major producers because of their petroleum deposits and qood sea
port facilities. Table 9-2 presents o geographical distribution of units by
reported capacity. Approximately 12 percent of these units produce fewer
than 5,000 Mg. Another 12 percent of the units have production capacities in
excess of 500,000 Mg. Seventy-five of these large facilities are located in
the southwest central region of the United States, which includes Texas and
Louisiana. Table 9-3 presents the total reported capacity for each region
by unit size.
*The 1976 version of the Organic Chemical Producers Data Base is used because
it was the most recent version available.
9-2
-------
TABLE 9-1.
State
Total
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawai i
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New York
North Carolina
Ohio
Oklahoma
Oregon
Pennsylvania
Puerto Rico
Rhode Island
South Carolina
Tennessee
Texas
Utah
Vermont
Virgin Islands
Virginia
Washington
West Virginia
Wisconsin
Wyoming
ESTIMATED ANNUAL PRODUCTION
BY STATE, 1976*
Number
of units
1,270
25
2
1
12
120
5
18
14
14
20
1
2
85
31
11
1
27
54
1
17
27
28
6
15
16
1
4
2
5
131
52
50
88
9
17
75
13
8
27
24
126
4
1
2
22
13
24
18
1
Percentage of
units reporting
capacity
40
52
50
00
33
23
20
26
50
43
30
00
50
31
26
64
00
56
74
00
35
48
25
17
47
37
100
100
50
40
24
19
38
26
78
53
27
85
00
48
46
76
25
100
100
68
46
63
22
100
"CAPACITY
Estimated total
capacity,
(103 Mg)
319,835
5,174
399
91
1,982
19,650
644
2,765
2,031
3,257
3,459
91
97
16,517
3,551
1,698
390
6,062
31,810
390
2,160
4,835
9,735
574
1,999
4,072
222
103
122
483
28,070
10,586
7,283
14,576
702
3,838
14.634
7,259
815
3,875
6,809
77,189
628
2
643
3,581
2,502
9,242
3,514
24
Capacities were estimated by calculating the mean of reported unit capacity
for each chemical. This was substituted for any missing values of unit
capacity for each chemical. If no units reported capacity for a chemical,
then the mean of all chemicals was substituted for the missing value. Esti-
.,;'•mated/.capacity.;.represents-the .sum.of reported capacities, means of-reported
capacity • fpr'j^'spineVspec i fie .chemicals, and Industry' mean reported capacity
for other chemicals.'
9-3
-------
TABLE 9-2
. DISTRIBUTION OF UNITS
BY UNIT CAPACITY AND REGION, 19764
Number of
units
Unit capacity ranges (103 Mg)
Region
North east
New England
Mid-Atlantic
North central
East
West
•p> South
East south
central
West south
central
South
Atlantic
West
Mountain
Pacific
0-5
17
7
10
16
13
3
23
2
4
17
3
. 0
3
5-10
5
0
5
5
3
2
7
3
3
1
2
1
1
10-25
" 12
4
8
13
10
3
34
12
10
12
17
3
14
25-50
16
5
11
13
9
4
38
8
18
12
12
2
10
50-100
17
4
13
13
9
4
44
5
20
19
5
0
5
100-250
10
0
10
17
16
1
50
9
24
17
8
0
8
250-500
5
1
4
6
6
0
34
3
24
7
3
0
3
500+
1
0
1
3
2
1
50
4
44
2
0
0
0
Units
reporting
capacity
83
21
62
86
68
18
280
46
147
87
50
6
44
Units not
reporting
capacity
235
39
196
202
182
20
200
45
54
101
119
10
109
Total
units
318
60
258
288
250
38,
480
91
201
188
169 •
16
153
Total
59
19
76
79
79
85
48
54
499
756
1,255
-------
\ _..,..^_.. TABLE 9-3.
'-"- ' " i
' i
I
i
t
1 Region
i
i
DISTRIBUTION OF
INDUSTRY CAPACITY BY UNIT
CAPACITY
AND REGION,
19764
i
i Industry capacity
\
i 1
i
1
1
— , 1 Is-""'"
lj North easjtf
; | New England
i Mid-Atlajfoic
,-J North cent-ral
East
West
South
-•.-
0-5
37.7
19.1
18.6
42.2
38.1"
4.1
62.9
5-10
43.1
--
43.1
38.8
22.5
16.3
58.8
10-25
199.2
70.8
128.4
205.9
142.4
63.5
605.6
25-50
548.4
176.9
371.5
478.1
350.2
127.9
1,442.0
Unit capacity
50-100
1,177.6
263.1
914.5
953.0
664.1
288.9
3,236.1
(103 Mg)
100-250
1,392.6
--
1,392.6
2,544.7
2,406.8
137.9
8,369.
150-500
1,811.2
299.4
1,511.8
2,024.9
2,024.9
— -
11,910.9
£500
742.5
--
742.5
2,430.8
1,906.9
523.9
67,600.5
Total
5,952
829
5,123
8,718
7,556
1,163
93,287 -
/[ East south
\o central
1.3
23.6
204.6
299.4
316.2
1,494.6
984.5
3,132.6
6,457 f
(In West south
central
South Atlantic
West
12.2
49.4
5.0
Mountain
Pacific
5.0
28.8
6.4
13.2
6.4
6.8
194.6
206.4
298.4
48.5
249.9
Caribbean
Total
147.8
153.9
1,309.1
710.8
431.8
477.2
72.6
404.6
109.8
3,055.5
1,532.3
1,387.6
342.5
--
342.5
72.6
5,781.8
4,411.7
2,463.5
1,055.1
--
1,055.1
--
13,362.2
8,596.6
2,329.8
1,024.2
--
1,024.2
1,324.1
18,095.3
63,197.8
1,270.1
--
--
—
5,659.1
76,432.9
78,685
8,145
3,216
127
3,088
7,166
118,339
-------
9.1 .3 Production and Sales
Production and sales data for the SOCMI are presented in Table 9-4. The
production of SOCMI chemicals increased from 58,050 Gg in 1968 to 84,530 Gg
in 1978, at an average annual growth rate of approximately 3.5 percent.
However, output levels have fluctuated widely since 1974. The effects of the
oil embargo, the increase in energy and feedstock prices, and the sharply
reduced demand resulting from a major economic recession caused the industry
to cut back production by 13.2 percent in 1975. In 1976 output rose only
slightly, but, in 1977, as real prices for energy and feedstocks fell, the
economy recovered, and the need to increase inventories became urgent, pro-
duction increased by 50.4 percent. In 1978 energy and feedstock prices began
to increase again and the need to replenish inventories disappeared. Output
declined that year by 28.6 percent. Nevertheless, production 1 978 was greater
than in 1974, suggesting that the industry may have substantially adjusted to.
the shocks experienced in 1974 and 1975.
Sales and production trends were virtually identical over the period
1967-1968*. The two variables are likely to remain highly correlated in the
future, because the industry's feedstock requirements are closely tied to its
production levels. The absolute level of sales was much lower than the level
of production (45.6 percent of production) over the period 1967 to 1978. The
difference between output and sales represents captive consumption, indicat-
ing that the indsutry has a relatively high degree of vertical integration.
9.1.4 Resource Use
Estimates of employment, assets, cost of materials and energy used in
SOCMI from 1972 to 1976 are presented in Table 9-5. In general, resource use
increased with production. Total industry employment, including administra-
tive, clerical, marketing and service employees as well as production workers,
increased 5 percent from 1972 to 1976. Employment of production workers
increased 4.1 percent during this period, although the number of production
workers declined during the adjustment period following the 1973-74 oil
*The estimated correlation coefficient for the two variables over this period
is 0.97. .
9-6
-------
TABLE 9-4. ANNUAL PRODUCTION WID~~SAtES OTTVNT HETIC~0RGAN'IG CHEMICALS5
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Production3
(Gg)
51,380
58,050
65,210
68,140
69,020
76,740
81,220
83,720
72,660
76,030
114,320
84,530
Sales volume9
(Gg)
23,440
26,960
30,360
32,090
33,020
36,930
39,420
38,450
32,920
32,520
49,470
35,310
Sales value9
($106)
3,085.55
3,411.91
3,590.07
3,702.20
3,724.03
4,173.9.7
4,991.53
9,357.99
8,411.34
10,187.76
15,317.72
12,951.16
Average
unit value3
($Ag)
0.13
0.13
0.11
0.12
0.11
0.11
0.13
0.24
0.26
0.29
0.31
0.36
See Appendix El for a discussion of the methodology used to compute these
data.
9-7
-------
TABLE 9-5. SOCMI RESOURCE USE
Year
1972
1973
1974
1975
1976
Total
employment
(103)
130.6
132.3
130.1
132.7
137.1
Production
workers
(103)
83.2
85.1
84.0
82.7
86.6
Total
assets
($1CP)
12,287.8
13,048.3
13,919.5
16,198.2
18,788.3
Cost of
materials
($106)
5,338.5
6,311.8
10,388
11,569
14,503.1
Energy
purchased
for heat
and power
(109 joules)
1,220.1s
l,286.6a
1.322.73
1,154.4s
1,202.4
These data were estimated by multiplying the 1976 estimate of energy use by
the ratio of production in each of the previous years to 1976 production
levels. Thus, for example, energy use in 1972 was estimated by multiplying
energy use in 1976 by the ratio of production in 1972 to production in 1976.
;9-8
-------
embargo. Cost of materials increased substantially during the 1972-1976
period; however, much of this increase can be attributed to rapidly increasing
prices of crude petroleum. Consumption of energy for heat and power has
fluctuated, resulting in an overall decrease of about 1 percent from 1972 to
1976. Value of assets increased each year from 1972 to 1976. The total
increase during that period was approximately 53 percent, much of which can
be accounted for by changes in the value of buildings and equipment. The
stock of physical assets increased at a much slower rate.
9.1.5 Consumption
The chemicals in SOCMI have a wide variety of end uses as fuels, solvents,
pesticides, and pigments, and as feedstocks for the production of plastics,
synthetic fibers and textiles, soaps and detergents, rubber products,
medicines and fertilizers. It is not possible to estimate consistently
apparent consumption, because import and export data presented in Table 9-6
for SOCMI are not compatible with the production and sales data presented
in Table 9-4. However, it is probable that historical consumption trends
have been similar to historical production and sales trends. Certainly, over
the period 1967-1978, consumption increased, although since 1974, if the sales
data presented in Table 9-4 can be regarded as an indicator of consumption,
consumption exhibited wide year-to-year variations for the reasons discussed
in Section 9.1.3.
9.1.6 prices
The general level of prices for SOCMI chemicals more than tripled
between 1967 and 1978. Most of the increase occurred after 1973. From 1967
to 1973, the average unit price of SOCMI chemicals remained close to $0.12/kg.
Following the 1973-1976 adjustments in oil prices, average prices in SOCMI
doubled, rising to $0.24/kg. After that time average unit prices increased
at a rate of approximately 11 percent annually, to a price of $0.36/kg in
1978. It is important to realize that these are average prices per unit of
all SOCMI chemicals. In 1976, prices for individual chemicals ranged from
$0.11/kg for formaldehyde to $4.30/kg for benzophenone. Changes in the unit
price for individual chemicals may vary substantially from the changes in
^fj;ay1era^e}ip'£it^ annual statistics of production, sales
;i. '.-• '"•••'".'•;-fif1i's;'i!('-..i"'; •"••."• ".'i1'1",1.';*0'' ''
volume, sales value and average unit value for the industry. The data are
9-9
-------
weighted using the procedures described in Appendix El to reflect the
behavior of the industry as accurately as possible.
9.1.7 International Trade
Chemical imports were first made subject to tariffs at the beginning
of the 20th century. The tariffs were initiated to protect the infant chemi-
cal industry from foreign competition. Since 1936, tariffs have been pro-
gressively lowered on chemical products. The U.S. International Trade
Commission reports 824 benzenoid intermediates on which tariffs are collected.
Of these, 179 are assessed duties competitively using import prices as the
basis for tariffs. Another 430 of these products are classified noncompeti-
tive, with tariffs based on U.S. domestic prices. The competitive status of
15 products is not available.8 The remainder are not tariffed.
Accurate data concerning imports and exports of SOCMI chemicals are not
available. The most reasonable approximation of trade statistics for SOCMI
are provided by the U.S. International Trade Comission. Annual value of
imports and exports for the period 1966-1967 is presented in Table 9-6. In
each of these years, U.S. exports exceeded U.S. imports of industrial organic
chemicals. Table 9-7 presents imports, exports and trade balance of indus-
trial organic chemicals in 1976 and 1977 between the United States and its
principal trading partners. These countries include West Germany, Italy,
the United Kingdom, Switzerland, France, Belgium, the Netherlands, Canada,
Japan, Mexico, Brazil, and Argentina. In 1977 the U.S. experienced a deficit
in its balance of trade in chemicals with West Germany, Japan, Italy, the
United Kingdom, Switzerland and France. It experienced a surplus in its
balance of trade in chemicals with Belgium, Canada, the Netherlands, Mexico,
Argentina and Brazil. Table 9-8 presents the value of imports for consumption
from principle sources from 1972 to 1977. These imports amounted to a total
of about $326 million in 1977.
9.1.8 Industry Growth
A number of forecasts of economic growth in the organic chemical in-
dustry are available. The annual growth rate used here, 5.9 percent, was
estimated by McGraw Hill10 for the basic organic chemicals industry. The
McGraw Hill estimate was selected for the following reasons. First, the
growth rate was calculated for a" group of chemicals which closely corresponds
9-10
-------
Year
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
TABLE 9-6. INDUSTRIAL ORGANIC CHEMICA
U.S. IMPORTS AND EXPORTS, 1966-77y
Imports
($106)
48
48
67
84
91
129
150
169
<5
259
205
294
326
LS:
Exports3
($106)
211
231
292
290
336
304
320
484
930
779
1,008
995
alncludes exports of some finished products. Figures include estimates
and are not strictly comparable with imports or production.
9-11
-------
TABLE 9-7. INDUSTRIAL ORGANIC CHEMICALS:
U.S. TRADE, BY PRINCIPAL TRADING PARTNERS, 1976 AND 1977 y
($103)
Source Imports Exports Trade balance
1976:
West Germany 94,768 10,487 - 84,281
Japan 61,228 27,380 - 33,848
Italy 30,678 N.A. - 30,000
United Kingdom 24,709 15,497 - 9,212
Switzerland 17,280 2,681 - 14,599
France 12,371 11,401 - 970
Belgium 2,154 46,779 44,625
Canada 8,081 93,471 85,390
Netherlands 8,987 178,111 169,124
Mexico 3,452 63,964 60,512
Argentina 1,927 N.A. - 1,500
Brazil 98 59,444 59,346
All other 28,103 498,985 470,882
Total 293,836 1,008,200 714,364
1977:
West Germany
Japan
Italy
United Kingdom
Switzerland
France
Belgium
Canada
Netherlands
Mexico
Argentina
Brazil
All Other
Total
o
105,172
65,770
32,711
31,132
21,956
15,763
9,839
7,270
4,858
4,673
3,353
538
22,865
325,900
5,038
30,736
N.A.C
27,458
6,541
N.A.C
61,126
82,676
156,581
62,965
6,283
78,512
477,469
995,385
-100,134
- 35,034
- 32,500
- 3,674
- 15,415
- 15,500
51,287
75,406
151,723
58,292
2,930
77,974
454,604
669,485
Data represent customs import value, the value appraised by the U.S. Customs
Service in accordance with the legal requirements of sec. 402 and 402a of
the Tariff Act of 1930, as amended.
Includes exports of some finished products. Figures include estimates and
are not strictly comparable with imports.
CN.A. = Not available.
9-12
-------
ORGANIC CHEMICALS: ' 9
IMPORTS FOR CONSUMPTION, BY PRINCIPAL SOURCES, 1972-77
Source
1972
1973
1974
1975
1976
1977
West Germany
Japan
Italy
United Kingdom
Switzerland
France
Belgium
Canada
Netherlands
Mexico
Argentina
All other
Total
66,085
36,181
11,305
7,605
11,593
1,611
1,220
4,301
5,067
35
3
5,031
150,037
72,715
29,793
10,705
10,433
16,063
4,233
7,919
5,515
4,724
486
--
6,892
169,478
84,059
65,027
17,323
21,119
15,846
8,585
10,494
4,826
10,291
1,812
--
19,190
258,572
62,145
49,243
19,073
18,820
14,773
9,797
1,871
4,352
6,738
388
657
17,625
205,482
94,768
61,228
30,678
24,709
17,280
12,371
2,154
8,081
8,987
3,452
1,927
28,201
293,836
105,172
65,770
32,711
31,132
21,956
15,763
9,839
7,270
4,858
4,673
3,353
23,403
325,900
Customs import value, the value appraised by the U.S. Customs Service in
accordance with the legal requirements of sec. 402 and 402a of the Tariff Act
of 1934, as amended.
9-13
-------
to the SOCMI chemicals. Second, the method used by McGraw Hill to develop
the growth rate is internally consistent and takes account of forecasted
developments in the U.S. economy. Third, the projections are developed for
the period 1976-1991, entirely covering the forecast period of interest in
this study (1981-1985).
In order to estimate the number of new model units covered by the regu-
latory alternatives, it is assumed that the number of operating facilities
will grow at the same rate as the industry's output. It is further assumed
that any regulatory alternative will take effect on January 1, 1981, and
therefore that the fifth year of the impact analysis is 1985. In 1976 (the
most recent year for which data are available), 1,334 facilities manufactured
SOCMI chemicals in the U.S. ' If the industry grows at an annual rate of 5.9
percent, by the beginning of 1981 this number will have risen to 1,678 faci-
lities and by the end of 1985 to 2,235 facilities. Thus, an estimated 557
units built to provide additional capacity for the industry will be covered
by the regulatory alternatives.
The regulatory alternatives will also cover units constructed to replace
existing capacity which "wears out" during the period. The number of replace-
ment units is estimated on the basis of the following assumptions. First,
units have a working life of 20 years. Second, the historical growth rate
for SOCMI prior to 1977 was 6 percent per year. Using these assumptions,
it is estimated that 274 new units will be required to replace the part of
the existing capacity that will "wear out" over the period 1981 to 1985.
The methodology used to compute this estimate is described in detail in
Appendix E2.
The estimates for entirely new units, combined with estimates for
replacement facilities, indicate a total of 831 units that will be affected
by the regulation. To estimate the number of A, B and C model units (identi-
fied in Section 6.1) that will be constructed between 1981 and 1985, it is
assumed that the mix of model units will not change over time, and that the
percentages of A, B and C model units are as follows:
Model unit Percent of existing units
A 52
B 33
C 15
9-14
-------
Number
of firms
Top 4
Top 8
Top 20
Top 40
TABLE 9-9.
Percent
of firms
0.72
1.43
3.58
7.17
INDUSTRY CONCENTRATION,
Estimated
capacity (gg)
58.75
91.82
145.75
186.68
19764
Percent of
industry capacity
18.3
28.6
45.4
58.1
9-15
-------
If these percentages are applied to the estimate of the total number of
units presented above, they imply that 432 A units, 274 B units and 125 C
units will be affected by the regulatory alternatives.
9.2 ECONOMIC IMPACT ANALYSIS
9.2.1 Market Structure and Financial Profile
SOCMI producers manufacture chemicals, each of which has its own
national and regional markets. Consequently, SOCMI firms encounter a wide
range of market situations for the different chemicals they produce. Many
SOCMI chemicals, for example, formaldehyde, urea and benzene, are manufac-
tured by a relatively large number of firms using an array of different
processes. The products have a wide range of end uses in which substitute
materials can often be used. Thus industry-wide elasticities of demand for
the chemicals are relatively high. In this type of market situation, pro-
ducers have little or no ability to pass on cost increases to consumers in the
form of higher market prices. Other SOCMI chemicals, for example, succino-
nitrile, isoamylene, and methyl butynol, are manufactured by a small number
of producers and in some cases only one producer, and have no close substi-
tutes in their end uses. In these oligopolistic and monopolistic markets,
producers may be able to exercise considerable influence on market prices
and to pass on a large part or all of any production cost increases in the
form of higher prices.
The ability oF firms to pass on cost increases in the form of price
increases is influenced by the extent to which the industry is vertically
and horizontally integrated. There is extensive vertical integration within
the SOCMI. Captive consumption in the industry averaged 53.7 percent* of
total output during the period 1967-1978, and this ratio varied only slightly
from year to year. The precise degree of horizontal integration within SOCMI
is difficult to evaluate because it varies considerably among products.
However, a general assessment of the industry-wide situation may be made
using the capacity share data presented in Table 9-9. These data suggest
*This figure is estimated from data presented in Table 9-4.
9-16
-------
that no one company or group of companies has a dominant position within the
industry. In 1976, the top four companies owned only 18.3 percent and the
top twenty firms 45.4 percent of total SOCMI capacity. There is no reason
to believe that the extent of industry-wide market concentration has altered
significantly since that time.
Data on the returns on equity, returns on debt, returns ,on preferred
stock, debt-asset ratios, equity-asset ratios and preferred stock-asset
ratios were collected for a sample of 100 chemical manufacturing firms for the
most recent available years.t These data are presented in Table E3-2. The
data have been used to estimate the cost of capital to firms in the SOCMI,
using the assumption that the sample of firms in Table E3-2 is unbiased and
normally distributed. A detailed discussion of the methodology used to
estimate the cost of capital is presented in Appendix E3.
The estimated cost of capital, presented in Table 9-10, is used in
Section 9.2.3 to estimate the economic impacts of SOCMI fugitive emissions
regulatory alternatives. Note that the average aftertax cost of capital for
chemical firms is 10.8 percent. On a pretax basis, this figure increases
to 20.8 percent. If, as was assumed, capital costs are normally distributed,
then 95 percent of the firms in the industry face aftertax costs of capital
in the range of 9.0 percent to 12.7 percent and pretax costs of capital in
the range of 17.2 percent to 24.4 percent.
9.2.2 Regulatory Alternatives
The four regulatory alternatives being considered are described in
detail in Section 6.2. The baseline regulatory alternative (alternative I)
does not require producers to implement additional control techniques. Con-
sequently, model units complying with this alternative would not incur any
incremental costs* and no economic impacts would result from its implemen-
tation. Regulatory alternatives II, III and IV require successively more
stringent equipment inspections and equipment specifications. Firms
lUata on the ratio variables and rates of return were available for 1977 and
1978, respectively.
incremental .costs of a regulatory alternative are those additional costs a
firm incurs in meeting the regulatory alternative that it would not incur
in meeting the baseline alternative.
9-17
-------
TABLE 9-10. ESTIMATED COST OF CAPITAL FOR FIRMS IN SOCMI3
Mean
Standard
deviation
Minimum
Maximum
Aftertax cost
of capital
10.807%
0.930
8.015%
12. 798%
Pretax cost
of capital
20.783%
1.789
15.414%
24.612%
See Appendix E3 for details of the data and methodology used to estimate
the cost of capital for firms in SOCMI.
9-18
-------
complying with regulatory alternatives II, III and IV would therefore incur
incremental costs, and consequently economic impacts would result from their
implementation.
9.2.3 Economic Methodology
9.2.3.1 Regulatory Scenarios. Economic impacts are estimated for
regulatory alternatives II, III and IV but not for regulatory alternative I,
since firms will not incur incremental costs in complying with that alter-
native. The economic impacts associated with alternatives II, III and IV
are estimated under two alternative assumptions about firm pricing behavior:
(1) full cost absorption and (2) full cost pricing. Combining the three
regulatory alternatives with the two alternative pricing models yields six
regulatory scenarios:
Regulatory Alternative Pricing Policy
Scenario 1 Alternative II Full Cost Absorption
Scenario 2 Alternative II Full Cost Pricing
Scenario 3 Alternative III Full Cost Absorption
Scenario 4 Alternative III Full Cost Pricing
Scenario 5 Alternative IV Full Cost Absorption
Scenario 6 Alternative IV Full Cost Pricing
Under full cost absorption, tho affcclod firm bears the full incremental
costs of environmental controls, ficceptimi a.lower rnl;e of return on its
9-19
-------
capital investment. Under full cost pricing, the firm adjusts product prices
so as to maintain its current aftertax rate of return on capital investment.
The alternative assumptions about firm pricing behavior are associated
with different market conditions in the affected industry. In both cases,
firms are assumed to have no monopsony power in resource markets. Thus, they
cannot pass back cost increases to resource suppliers. In the cost absorption
case, the domestic industry as a whole is assumed to be a price taker, unable
to affect the market price of its product either because of the existence of
close product substitutes, or because of strong international competition in
domestic and foreign markets. However, full cost pricing will take place if
the industry produces a commodity for which no domestic or imported substitutes
exist, or if the industry has constant costs. A constant-cost industry is
one in which unit costs remain constant as industry output increases. Firms
in such industries experience constant returns to scale.
In fact, firms in SOCMI face a wide variety of product market situations
(see Section 9.2.1). Some firms will be able to fully pass through cost
increases to consumers in the form of higher prices. Some will be able to
pass on only a part of the c'ost increases. Others will be forced to fully
absorb all regulatory control costs, leaving product prices unchanged. Conse-
quently, the full cost pass through and full cost absorption scenarios
evaluated below provide estimates of the maximum range of possible price and
rate of return impacts for the different products and firms in SOCMI.
9.2.3.2 Estimation of Regulatory Price Impacts Under Full Cost Pricing.
Under full cost pricing, the firm is assumed to respond to cost increases by
adjusting product price to maintain a target rate of return on investment.
The required price change (dP) may be calculated using the following
equation:*
dp _ dTQP + r dK/Q-t) (])
Q
where dP = required change in product price
dTOC = total annual operating costs of compliance
*The derivations of Equations (1) and (2) are presented in Appendix E4.
9-20
-------
dk = total initial costs of compliance
Q = total annual unit output
r = target rate of return
t = tax rate
9.2.3.3 Estimation of Rate of Return Impacts Under Full Cost Absorption.
Under full cost absorption, an increase in facility production costs results
in a lower rate of return on investment for the firm, because market condi-
tions prohibit it from passing on cost increases to the consumer. The impact
on the facility's rate of return on investment is given by the following
equation:
_dr = r ' dK + (1-t) dTOC (2)
K
where dr = change in rate of return, and
K = preregulation level of capital investment.
Note that pretax rate of return impacts may be calculated by setting the tax
rate variable, t, equal to zero in Equations (1) and (2). Also note that
price and rate of return impacts are estimated on the assumption that capacity
utilization rates remain constant (that is, Q remains unchanged). To the
extent that the regulatory alternatives result in decreases (increases) in
capacity utilization rates, price and rate of return impacts will be larger
(smaller) than those estimated using Equations (1) and (2) because of
economies of scale in the use of control techniques.
9.2.3.4 Other Economic Impacts. The price and rate of return impacts
estimated by the above techniques are used to make a quantitative assessment
of the probable impacts of the regulatory alternatives II, III, and IV on
industry growth, new facility openings, the replacement of existing facilities,
and investment levels. These data are then used to assess the extent of
interindustry and macroeconomic impacts associated with the various regulatory
alternatives.
9.2.3.5 Estimation Data. Estimation of price and rate of return impacts
for different model units requires data on the following variables: (1) total
acquisition and installation costs of the control equipment (dK), (2) total
annual operating costs of the control equipment and monitoring procedures
(dTOC), (3) the preregulation capital stock (k), (4) the target rate of
9-21
-------
return (r), (5) the tax rate (t), and (6) model unit production levels (Q).
Data on dK and dTOC for each of the three model units identified as
representative of the industry* were obtained from Section 8.1 The tax rate
is assumed to be 48 percent. Data on model unit production levels were
obtained from the 1976 Organic Chemical Producers Data Base for each of the
units covered by the regulatory alternatives. The 831 model units are
assumed to be distributed by capacity in an identical manner to the 1,105
units for which both value of product and quantity data are available in the
1976 Organic Chemical Producers Data Base. Thus, the number of new units with
a given capacity, say 100 Gg, is assumed to be equal to the number of units
in the data base (831/1,105). Actual unit output levels are obtained by
applying a capacity utilization rate to the estimated unit capacities.
To evaluate industry-wide impacts, the cost data from Section 8.1 were
adjusted to allow for higher or lower product recovery credits for chemicals
with a value greater or less than $0.36/kg. For such chemicals, product
recovery credits were estimated by multiplying estimated product savings by
the price of the chemical in question.
Data on the value of the preregulation capital stock for plants of
different capacities were calculated as follows: A capital-capacity coef-
ficient for firms in SOCMI was obtained by dividing the estimated total
value of industry assets in 1976 by the volume of output produced in that
year.** The estimate of the capital-output coefficient was converted into a
capital-capacity coefficient by multiplying the capital-output coefficient
by an assumed industry-wide capacity utilization rate. The assumed capacity
utilization rate for 1976 was 50 percent. This capacity utilization estimate
was based on the assumption that the typical capacity utilization for the
industry is 75 percent. In 1976, output was 9.2 percent below the industry-
wide high level of output achieved in 1974. Between 1974 and 1976 it is
probable that some additions to industry capacity were made. Hence, the
assumption of a 50 percent capacity utilization rate for 1976, though somewhat
*See Chapter 6 for a detailed discussion of the model units.
**See Tables ;9-4 and 9-5 for data on production and total industry assets.
9-22
-------
arbitrary, is not unreasonable. The 1976 capital-capacity coefficient was
updated to last quarter 1978 dollars using the machinery and equipment price
index computed by the United States Department of Commerce.^ The capital -
capacity coefficient estimated by the above procedure was $125/Mg of product.
This coefficient was multiplied by model unit capacity to obtain an estimate
of K for each model unit considered in the analysis.
Estimates of pretax and post-tax rates of return used in the analysis
are presented in Table 9-10. These data were obtained from an analysis of
a sample of 100 firms in the SOCMI industry. Details of the analysis are
contained in Appendix E3.
9.2.4 Economic Impacts
9.2.4.1 Rate of Return Impacts Data on unit capacity, product value,
capital investment and tax rates are available for 1,105 units in the 1976
Organic Chemical Producers Data Base. Price data were updated using the
Chemical Marketing Reporter. Capital stock estimates were also expressed
in 1978 prices. These data were used in conjunction with the cost information
presented in Section 8.1 to calculate full cost absorption rate of return
impacts of regulatory alternatives II, III, and IV for the 831 model units
projected to be built. It is assumed in estimating the rate of return
impacts presented here that the 831 new model units will have the same
capacity and product value distributions as the units in the Orqanic Chemical
Producers Data Base, and that capacity utilization for each unit, is 50 percent.
This relatively low capacity utilization rate is used to estimate unit output
levels because it represents a feasible worst-case economic scenario for the
industry (that is, economic conditions similar to. those experienced in 1976).
As a result, actual impacts are likely to be less adverse than those presented
below.
Rate of return impacts are estimated on the basis of these assumptions
for each of the 831 new model units covered by the regulatory alternatives.
It is probable that the assumption of a constant product price distribution
also results in an overestimate of adverse rate of return and price impacts,
since the prices of the products manufactured by SOCMI are expected to in-
crease between 1979 and 1985 as energy and feedstock costs rise. Any real
9-23
-------
increase in product prices will raise the value of product recovery credits,
lower the net costs of compliance associated with any given regulatory Alter-
native, and thereby reduce adverse rate of return and price impacts.
Rate of return impacts for A, B, and C average model units under each
regulatory alternative are presented in Table 9-11. Each of these average
model units is assumed to manufacture products valued at approximately
$0.36/kg, to have an annual capacity of 84,678 Mg and to have an existing
cost of capital of 10.81 percent. These average model units differ only in
terms of the complexity of the processes they use to manufacture the chemicals.
The product value and rate of return data represent the means for each
variable in the samples used in the analysis. Under regulatory alternatives
II and III, each average model unit experiences a very small increase, not
a decrease, in its aftertax rate of return on investment, regardless of the
process it uses. This result is obtained because at a price of $0.36/kg for
recovered product, product recovery credits exceed total annualized costs
of control. Under regulatory alternative IV, average model unit of types A
and B experience small decreases in aftertax rates of return on investment.
Model C units experience rate of return decreases amounting to 1.12 per-
centage points, still quite small adverse impacts.
The data presented in Table 9-11 suggest that some firms may benefit
from the implementation of any regulatory alternative. These results are
subject to the following qualifications. In the above analysis, it is assumed
that firms will not independently implement the emissions controls proposed
in the regulatory alternatives. In fact, if there are significant net cost
reductions to be achieved from additional emissions controls, firms will
voluntarily adopt them. Under such circumstances, the cost reductions associ-
ated with any regulatory alternative will be considerably smaller. Note that
incentives for voluntary emissions controls increase as the value of the
manufactured product increases. As some SOCMI producers manufacture highly
valued products with prices in excess of $0.50/kg, they are likely to be
willing to use extensive emissions control techniques in the absence of any
NSPS.
9-24
-------
TABLE 9-11. AVERAGE RATE OF RETURN IMPACTS3
Model
Units
Unit A
Unit B
Unit C
Change in
Alternative II
+0.000
+0.003
+0.006
rate of return
Alternative
+0.000
+0.001
+0.005
(percentage points)
III Alternative IV
-0.16
-0.37
-1.12
Impacts are estimated on the assumption that the initial aftertax rate of return
on investment is 10.807 percent, the mean cost of capital presented in Table 9-10;
the initial price of the product is $0.36/kg; plant capacity is 84,678 Mg; and
the capacity uti1ization rate is 50 percent.
9-25
-------
Although in general firms will not be affected by the implementation of
regulatory alternatives II, III, and IV, because of wide variation in produc-
tive capacity and value of product among model units, some will experience
adverse rate of return impacts. Estimates of the numbers of model units
experiencing rate of return decreases in excess of one and two percentage
points as a result of the implementation of each regulatory alternative are
presented in Table 9-12. These estimates were obtained by calculating rate
of return impacts for each of the 831 new model units under the assumption
that 52 percent, 33 percent, and 15 percent of all units of all sizes are
A, B and C model units, respectively. Under regulatory alternatives II and
III, the estimated number of adversely affected units is very small; only 6
and 12 units, out of a total of 831 model units, experience rate of return
decreases of more than one percentage point. Under regulatory alternative IV,
a much larger number of units, 93 in all, are estimated to experience rate
of return decreases in excess of 1 percent under a full cost absorption
scenario. It should be noted that most of these adversely affected units
are B and C model units rather than A model units. In fact, all B and C
model units with capacities in excess of 26,464 and 89,121 Mg, respectively,
producing chemicals with prices exceeding $0.15/kg, will experience rate of
return impacts smaller than one percentage point even under alternative IV.
The EPA estimates that virtually all B and C model units do in fact have
capacities in excess of this figure, and furthermore, industry sources indi-
cate that most produce chemicals that have prices in excess of $0.15/kg.13
If the estimated impacts on B and C model units are ignored, only 25 units
are likely to be adversely affected by regulatory alternative IV.
9.2.4.2 Price Impacts. The potential price impacts of regulatory
alternatives II, III, and IV are also estimated under the assumption that
capacity and value of product distributions will remain constant over the
forecast period, 1979-1985. The price impact estimates are therefore subject
to the same limitations as the rate of return impact estimates discussed
above. Potential price impacts for A, B and C model units with average
capacities of 84,678 Mg and product values of $0.36/kg are presented in
Table 9-13. Under regulatory alternatives II and III, price impacts are
9-26
-------
TABLE 9-12. MODEL UNITS EXPERIENCING SIGNIFICANT RATE OF RETURN IMPACTS
UNDER FULL COST ABSORPTION3
Model
units
Unit A
Unit B
Unit C
Total
Alternative
dr < -1% dr
6
0
0
6
II
< -2%
4
0
0
4
Alternative
dr < -1% dr
7
4
1
12
III
< -2%
5
2
0
7
Alternative
dr < -1% dr
25
34
34
93
IV
< -2%
12
16
20
48
dr denotes the percentage point change in firms' rates of return on investment.
TABLE 9-13. AVERAGE PERCENTAGE PRICE IMPACTS OF REGULATORY ALTERNATIVES3
Model
units
Unit A
Unit B
Unit C
Alternative II
-0.000
-0.002
-0.009
Price changes (percent)
Alternative III
-0.000
-0.002
-0.007
Alternative IV
+0.000
+0.000
+0.733
Impacts are estimated on the assumption that the target rate of return is
10.807 percent, the average cost of capital presented in Table 9-10; the
initial price of the product is $0.36/kg; plant capacity is 84,678 Mg; and
the capacity utilization rate is 50 percent.
9-27
-------
negative for each type of model unit because annual product recovery credits
exceed the total annualized cost of the monitoring procedures and capital
equipment required under these alternatives. Under regulatory alternative IV,
extremely small positive price impacts occur. In general most units will
not increase product prices as a result of the implementation of regulatory
alternatives II, III or IV. However, because of the variations in capacity
and product value within the industry, some firms may have to raise product
prices in order to maintain existing rates of return on investment. In some
cases, the price increases required by individual facilities are in excess of
5 percent and even 6 percent of the current product price. Data on the esti-
mated numbers of such units are presented in Table 9-14. These estimates are
also obtained by calculating price impacts for each of the 831 new model units
under the assumption that units are distributed among A, B, and C model units
in the manner described above and operate at 50 percent of unit capacity.
Under alternatives II and III, only five A and eight B model units would have
to increase product prices by more than 5 percent. Under alternative IV, 30
units must increase prices by more than 5 percent preregulation rates of
return on investment. However, it should be noted that these estimates may
overstate the extent of significant price impacts under regulatory alterna-
tive IV. Most of the units estimated to require price increases in excess
of 5 percent are C model units. In fact, C model units that manufacture
chemicals with prices in excess of $0.15/kg and have capacities greater than
71,550 Mg do not have to increase product prices by more than 5 percent to
maintain their target rates of return on investment. The EPA estimates that
virtually all plants using processes with the same degree of complexity as
that assumed for C model units have larger capacities and produce products
with higher values than these. In addition, model units are assumed to
operate at the relatively low capacity utilization rate of 50 percent.
A final caveat concerning pnice and rate of return impacts should be
noted. The impact estimates presented in Tables 9-12 and 9-14 were developed
on the assumption that feedstock prices are unaffected by the implementation
of any regulatory alternative. However, as the industry extensively uses its
own products as feedstocks, this assumption is not strictly valid and
9-28
-------
TABLE 9-14. MODEL UNITS REQUIRING SIGNIFICANT PRICE INCREASES
TO MAINTAIN TARGET RATES OF RETURN
Model
Units
Unit A
Unit B
Unit C
Total
Alternative II
Price increase
4
0
0
4
4
0
0
4
Alternative III
Price increase
fe D/D =; O/D
4
2
0
6
4
2
0
6
Alternative IV
Price increase
6
10
14
30
5
9
12
26
9-29
-------
introduces a systematic upward bias in the estimated size of adverse rate of
return and potential price impacts. The upward bias occurs because, in
general, firms adopting alternative II, III and IV control technologies will
achieve net cost reductions and, at least in competitive markets, will tend
to reduce rather than increase the prices of products used as feedstocks by
the industry.
9.2.4.3 Investment Impacts. It is difficult to assess the impact of any
of the standards on the number of units to be constructed between January 1,
1981, and December 31, 1985, because of the variations in these impacts across
units. Some smaller facilities may not be erected as a result of the standard
because of adverse impacts on rates of return and price competitiveness.
Other larger facilities may be built because production costs fall as a result
of emissions reductions and product recovery credits. Therefore, in this
analysis it is assumed that implementation of regulatory alternatives II, III,
and IV will have no measurable impact on the number of new facilities con-
structed between 1981 and 1985, the 5-year period following proposal of any
regulatory alternative. Industry-wide investment impacts are therefore simply
the incremental capital costs associated with the acquisition of the capital
and monitoring equipment required under each regulatory alternative by the
831 new units expected to be constructed between 1981 and 1985.
Data on these investment impacts are presented in Table 9-15. The esti-
mates are obtained by assuming that 432 A model plants, 274 B model units,
and 125 C model units will be constructed and that, as a result of each
regulatory alternative, these units incur incremental capital costs equal to
those presented in Section 8.2. Under regulatory alternatives II and III,
industry-wide investment impacts are quite small, less than $22 million.
Under regulatory alternative IV, they increase substantially to almost $233
million. Nevertheless, even under regulatory alternative IV, the total 5-year
investment impacts of any of the regulatory alternatives would be less than
1.14 percent of total industry assets in 1976 and less than 1.64 percent of
the value of industry sales in 1978.* It appears, therefore, that the
*Data on total industry value of assets and industry sales are presented in
Tables 9-4 and 9-5.
9-30
-------
TABLE 9-15. INVESTMENT IMPACTS
Model
units
Number of
model units
Incremental
model unit
costs of control
C$106)
Incremental industry
costs of control
($1CP)
Regulatory
alternative
Unit A
Unit B
Unit C
II
Total
Total
432
274
125
831
0.0132
0.0272
0.0660
831
Total
831
5.7
7.5
8.3
21.5
Regulatory
alternative III
Unit A
Unit B
Unit C
432
274
125
0.0132
0.0272
0.0660
5.7
7.5
8.3
21.5
Regulatory
alternative IV
Unit A
Unit B
Unit C
432
274
125
0.0871
0.2950
0.9110
37.6
80.8
113.9
232.3
9-31
-------
industry as a whole will not have much difficulty in obtaining the investment
funds to acquire required control equipment under any of the regulatory
alternatives.
9.2.4.4 Employment Impacts. Regulatory alternatives II, III, and IV
will each have small but measurable impacts on employment in SOCMI because
they require firms to intensify monitoring and maintenance schedules to
control fugitive emissions. Estimates of the number of additional workers
required as a result of each regulatory alternative are presented in Table 9-16.
The estimates were obtained by multiplying the projected numbers of each type
of affected facility by the unit-by-unit, person-year monitoring and mainte-
nance requirements for each standard presented in section 8.1* The largest
employment impacts (400 workers) are associated with regulatory alternative
III, which requires more stringent monitoring programs than alternative II.
Under alternative IV, some alternative III monitoring requirements are
replaced by equipment controls, reducing incremental employment requirements
to approximately 225 workers. The employment impacts of each of the standards
are small relative to total employment in the industry, representing no more
than 0.6 percent of the 1976 SOCMI work force in each case.
9.2.4.5 Total Annualized Costs of Control. Total incremental annualized
costs of control for the fifth year following promulgation of alternatives II,
III, or IV are presented in Table 9-17. Product recovery credits are calcu-
lated using the fourth quarter 1978 industry-wide average product price of
$0.36/kg. Under regulatory alternatives II and III, the industry as a whole
is estimated to reduce annualized production costs by $28.73 million and
$21.35 million, respectively. Under regulatory alternative IV, annualized
production costs are estimated to increase by $11.17 million. If the above
estimates are accurate in the minimal sense that they indicate the direction
in which production costs will move and their approximate order of magnitude,
then it may be concluded that none of the regulatory alternatives will result
in any measurable industry-wide increase in prices.
*A person-year is assumed to consist of 2,000 person-hours.
9-32
-------
All units
TABLE 9-16. EMPLOYMENT IMPACTS
(Person-years)
Model
unit
Unit A
Unit B
Unit C
Alternative II Alternative III Alternative IV
Unit Industry Unit Industry Unit Industry
0.1237 53.43 0.1855 80.14 0.0579 25.01
0.1863 51.05 0.5079 139.16 0.2277 62.39
0.5017 62.71 1.4532 181.65 1.0982 137.27
167.19
400.95
224.67
9-33
-------
TABLE 9-17. MODEL UNIT AND INDUSTRY ANNUALIZED CONTROL COSTS
Regulatory
alternative
Alternative II
Unit A
Unit B
Unit C
No. of
model
units
432
274
125
Incremental
unit
annual ized costs
without product
recovery. credit
C$103 )
12.1
20.3
48.3
Incremental
unit
annual ized costs
with product
recovery ^credit
($10 )
- 3.3
- 38.5
-134.0
Incremental
industry
annual ized costs
with product
recoveryocredit
($ioj)
- 1,430
-10,550
-16,750
Total
Alternative III
Unit A
Unit B
Unit C
Total
Alternative
Unit A
Unit B
Unit C
IV
432
274
125
432
274
125
15.0
35.5
95.1
29.0
93.5
283.0
- 1.7
- 28.3
-103.0
7.9
13.3
33.0
Total
-28,730
- 730
- 7,750
-12,870
-21,350
3,410
3,640
4,120
11,170
Product recovery credits estimated on the basis of an assumed product value of
$0.36/kg.
9-34
-------
9.2.4.6 Interindustry Impacts. Interindustry impacts will be negligible,
because net annualized costs of control are extremely small relative to the
value of total industry output, representing less than 0.03 percent of the
value of 1978 output in even the most adverse case (regulatory alternative IV).
9.3 SOCIO-ECONOMIC AND INFLATIONARY IMPACTS
The socio-economic and inflationary impacts of alternatives II, III and
IV will be very small.
(1) Annualized Costs: In the fifth year following promulgation, the
regulatory alternatives, if implemented, are estimated to result in either
annualized cost reductions or very small annualized cost increases. Conse-
quently, none of the alternatives violates the regulatory criterion of
$100 million.
(2) Price Impacts: Because industry-wide annualized costs of compliance
for alternatives II, III and IV are estimated to be negative or extremely '
small relative to the value of industry output, none of the standards is
likely to cause any industry-wide price increases.
9-35
-------
9.4 REFERENCES
1. Condensed Chemical Dictionary. Rev. by Gessner Hawley, 8th ed. New
York: Van Nostrand Reinhold Co, 1971.
2. Comer, James F. Synthetic Organic Chemicals Manufacturing Industry:
. Inputs and Product Uses. Prepared for the U.S. Environmental Protection
Agency, 1979.
3. Proceedings of the Conference on Chemical Feedstock Alternatives.
American Institute of Chemical Engineers, Houston, TX, 1977.
4. Radian Corp. "Organic Chemical Producers Data Base, 1976." Prepared
for the U.S. Environmental Protection Agency under EPA Contract No.
68-03-2623. 1978.
5. United States International Trade Commission. Synthetic Organic Chemi-
cals, U.S. Production and Trade. Washington, D.C^1967-1978.
6. United States Department of Commerce. Annual Survey of Manufactures,
Industry Profiles, 1976. Washington, D.C.
7. Russell, T.W. F. , M.W. Swartzlander, and J. Wei. The Structure of the
Chemical Processing Industries. New York: McGraw Hill, 1979. pp.
321-334.
8. United States International Trade Commission. Imports of Benzenoid
Chemicals and Products, 1978. Washington, D.C., July 1979.
9. United States International Trade Commission. "Import Penetration of
U.S. Markets for Cyclic Intermediates." Synthetic Organic Chemicals:
U.S. Production and Sales -1977. Washington, D.C., 1978.
10. The American Economy, Prospects for Growth to 1991. New York: McGraw
Hill, 1977.
11. Radian Corp. "Organic Chemical Producers Data Base, 1978." Prepared
for the U.S. Environmental Protection' Agency under EPA Contract No.
68-03-2623. 1979.
12. United States Department of Commerce. Survey of Current Business.
Washington, D.C., 1976-1979.
13. Chemical Marketing Reporter. September 7, 1979.
14. Memo from Hustvedt, K.C. November 7, 1980. Model Unit Capacities.
9-36
-------
APPENDIX A
EVOLUTION OF THE PROPOSED STANDARDS
Date
December 1978
3-5 January 1979
2 February 1979
16 February 1979
27 February 1979
5-8 March 1979
9 March 1979
14 March 1979
21 March 1979'
Action
Work began on developing standards for new
sources in SOCMI.
Testing at Stauffer Chemical Company in
Louisville, Kentucky (SOCMI Unit C).
Letter to Stauffer Chemical Company requesting
information pertaining to testing-at SOCMI
Unit C. .
Section 114 letter to Phillips Petroleum Company
requesting permission to perform emission
sampling of plant equipment.
Letter to Phillips Petroleum Company requesting
information on plant's directed maintenance
program.
Testing at Phillips Petroleum in Sweeny, Texas
(SOCMI Unit D).
Letter to Exxon Chemical Company requesting
information on fugitive emissions from cyclohexane
unit.
Comments requested from industry on Hydroscience
Draft Fugitive Emissions Report.
Letter from Exxon Chemical Company U.S.A.
Response to request for information on fugitive
emissions from cyclohexane unit.
-------
•f
Date Action
10 April 1979 Letter from Exxon Chemical Company USA.
Comments on Hydroscience draft, "Fugitive
Emissions Report," Feb. 1979.
10 April 1979 Letter from Tennessee Eastman Company.
Comments on Hydroscience draft "Fugitive
Emission Report," Feb. 1979.
Letter from Phillips Petroleum Company.
Comments on Hydroscience draft "Fugitive
Emission Report," Feb. 1979.
12 April 1979 Letter from Shell Oil Company. Comments
on Hydroscience draft Fugitive Emissions
Report, Feb. 1979.
12 April 1979 Letter from Vulcan Materials Company.
Review of the Hydroscience draft Fugitive
Emission Report.
3 May 1979 Letter from American Cyanamid Company.
Comments on Hydroscience Draft "Fugitive
Emissions Report," Feb. 1979.
8 May 1979 Letter from B.F. Goodrich Company. Comments
on Hydroscience draft "Fugitive Emissions
Report," Feb. 1979.
17 May 1979 Letter from Texas Chemical Council. Comments
on Hydroscience draft "Fugitive Emissions
Report," Feb. 1979
1 June 1979 Letter from Atlantic Richfield Company.
Comments on Hydroscience draft "Fugitive
Emissions Report." Feb. 1979.
-------
CJ
Date
12 June 1979
20 June 1979
21 June 1979
21 June 1979
21 June 1979
19 July 1979
17 October 1979
24 October 1979
24 October 1979
7 November 1979
12 November 1979
Action
Discussion of fugitive emissions sampling at
DuPont Chemical plants.
Meeting with California Air Resources Board in
Sacramento, California. Discussions of fugitive
emissions and regulations.
Meeting with ARCO in Carson, California.
Discussion of fugitive emissions and regulations.
Meeting with Chevron in El Segundo, California.
Discussion of fugitive emissions and regulations.
Meeting with South Coast Air Quality Maintenance
District in El Monte, California. Discussion of
fugitive emissions and regulation.
Chemical Manufacturers Association/Texas Chemical
Council Fugitive Emission Seminar, Washington,
D.C.
Letter from Century Systems Corporation. Cost
data for portable VOC detection instrument.
Letter to Exxon Chemical Company requesting
information on leak-free technology.
Letter to Dow Chemical U.S.A. requesting information
on leak free technology.
Chapters 3-6 of Background Information Document
sent out for public review.
Letter from Chemical Manufacturers Association.
Comments on Hydroscience draft "Fugitive Emissions
Report," Feb. 1979.
-------
I
-is.
Date
3 January 1980
4 January 1980
10 January 1980
5. February 1980
12 February 1980
2 April 1980
Action
Letter from Shell Oil Company. Comments
on draft BID sections, "Fugitive Emission
Sources in the Synthetic Organic Chemicals
Manufacturing Industry," Nov. 1979.
Letter from Phillips Petroleum Company.
Comments on draft BID sections, "Fugitive
Emission Sources in the Synthetic Organic
Chemicals Manufacturing Industry," Nov.
1979.
Letter from Vulcan Materials Company.
Comments on draft BID sections, "Fugitive
Emissions Sources in the Synthetic Organic
Chemicals Manufacturing Industry." Nov. 1979.
Letter from 3M. Comments on draft BID
sections, "Fugitive Emission Sources in
Synthetic Organic Chemicals Manufacturing
Industry," Nov. 1979.
Letter from Chemical Manufacturers Association.
Comments on Leak-Free Technology for Control
of Benzene Fugitive Emissions.
Meeting with CMA in Durham, North Carolina
to discuss recommended standard.
-------
Date Action
16 April 1980 Recommended Standard presented at
NAPCTAC meeting in Raleigh, North
Carolina.
21 April 1980 Letter from Colt Industries. Comments
on Selection of Packing.
23 April 1980 Letter from South Coast Air Quality
Management District. Comments about
recommended rules.
28 April 1980 Letter from American Cyanamid Company.
Comments on draft regulations discussed
at April 16-17 NAPCTAC meeting.
28 April 1980 Letter from 3M Corporation. Comments on
draft regulations discussed at April 16-17
NAPCTAC meeting.
29 April 1980 Letter from Oxirane Corporation. Comments
concerning draft SOCMI regulations.
1 May 1980 Telephone discussion with Hartford Steam
Boiler Insurance and Inspection Company
engineering department about use of rupture
disks and relief valves.
1 May 1980 Telephone discussion with Brown & Root, Inc.,
about use of rupture disks and relief valves.
2 May 1980 Letter from Brown and Root. Information
concerning relief devices.
20 May 1980 Meeting with Furmanite to discuss valve
repairability.
-------
Date Action
27 May 1980 Letter from Chemical Manufacturers
Association. Comments on SOCMI
regulations.
28 May 1980 Letter to A.W. Chesterton. Request
for information on pump seal performance.
3 June 1980 Letter from L. Bentsen, U.S. Senate.
Texas Chemical Council Comments on
development of SOCMI standard.
4 June 1980 Letter from J. Brooks, U.S. House of
Representatives. Texas Chemical Council
Comments on development of SOCMI standard.
5 June 1980 Letter from B. Eckhardt, U.S. House of
Representatives. Comments on development
of SOCMI standard by The Upjohn Company.
5 June 1980 Letter from J. Tower, U.S. Senate.
Comments on development of SOCMI standard
by the Upjohn Company.
12 June 1980 - Letter from Chemical Manufacturers Association.
Draft comments on development of SOCMI fugitive
standard.
13 June 1980 Letter from Chemical Manufacturers Association.
Comments on draft Hydroscience report.
16 June 1980 Meeting with DuPont in Durham, North
Carolina. Discussion of skip period
monitoring.
17 June 1980 Meeting with Chemical Manufacturers Association/
Texas Chemical Council. Discussion of draft
SOCMI regulations.
-------
Date Action
30 June 1930 Letter from Texas Chemical Council.
Comments on Draft BID and recommended
SOCMI standard.
18 July 1980 Meeting with Texas Chemical Council
in Durham, Morth Carolina. Discussion
of Draft BID and recommended standard.
28 July 1980 Letter from Texas Chemical Council.
Information concerning "capital creep."
18 August 1980 Letter from UOP. Questions about draft
regulations.
-------
-------
APPENDIX.B
INDEX TO ENVIRONMENTAL CONSIDERATIONS
This appendix consists of a reference system which is cross
indexed with the October 21, 1974, Federal Register (39 FR 37419)
containing EPA guidelines for the preparation of Environmental
Impact Statements. This index can be used to identify sections of
the document which contain data and information germane to any portion
of the Federal Register guidelines.
B-l
-------
TABLE B-l. INDEX TO ENVIRONMENTAL CONSIDERATIONS
CO
I
Agency Guideline for Preparing Regulatory
Action Environmental Impact Statements
(39 FR 37419)
(1) Background and summary of regulatory
alternatives
Regulatory alternatives
Statutory basis for proposing standards
Affected industry
Affected sources
Availability of control technology
Location Within the Background Information Document
The regulatory alternatives are summarized in
Chapter 1, Section 1.1, pages 1-1 through 1-4.
The statutory basis for the proposed standards
is summarized in Chapter 2, Section 2.1, pages
2-1 through 2-5.
A discussion of the industry affected by the
regulatory alternatives is presented in Chapter 3,
Section 3.1, pages 3-1 through 3-3. The industry
is further defined in Appendix F. Details of the
"business/economic" nature of the industry are
presented in Chapter 9, pages 9-1 through 9-35.
A description of the sources affected by the
regulatory alternatives is presented in Chapter 3,
Section 3.2, pages 3-3 through 3-17.
A discussion of available emission control
techniques is presented in Chapter 4, Sections
4.1 and 4.2, pages 4-1 through 4-24.
-------
TABLE B-l. (CONTINUED)
OD
I
Agency Guideline for Preparing Regulatory
Action Environmental Impact Statements
(39 FR 37419)
(2) Environmental, Energy, and Economic
Impacts of Regulatory Alternatives
Regulatory alternatives
Environmental impacts
Energy impacts
Cost impacts
Economic impacts
Locations Within the Background Information Document
Various regulatory alternatives are discussed in
Chapter 6, Section 6.2, pages 6-4 through 6-7.
The environmental impacts of the various regulatory
alternatives are presented in Chapter 7, Sections
7.1, 7.2, and 7.3, pages 7-1 through 7-12.
The energy impacts of the various regulatory
alternatives are discussed in Chapter 7,
Section 7.4, pages 7-12 through 7-13.
Cost impacts of the various regulatory alternatives
are discussed in Chapter 8, pages 8-1 through 8-27.
The economic impacts of the various regulatory
alternatives are presented in Chapter 9, pages
9-1 through 9-35.
-------
TABLE B-l. (CONTINUED)
DO
I
Agency Guideline for Preparing Regulatory
Action Environmental Impact Statements
(39 FR 37419)
(3) Environmental impact of the
regulatory alternatives
Air pollution
Water pollution
Solid waste disposal
Location Within the Background Information Document
The impact of the proposed standards on air
pollution is presented in Chapter 7, Section
7.1, pages 7-1 through 7-8.
The impact of the proposed standards on water
pollution is presented in Chapter 7, Section
7.2, pages 7-8 through 7-11.
The impact of the proposed standards on
solid waste disposal is presented in Chapter
7, Section 7.3, pages 7-11 through 7-12.
-------
APPENDIX C
EMISSION SOURCE TEST DATA
The purpose of Appendix C is to describe testing results used in
the development of the Background Information Document (BID) for fugitive
emissions from the Synthetic Organic Chemicals Manufacturing Industry
(SOCMI). The information in this appendix consists of a description of
the tested facilities, and the sampling procedures and test results of
fugitive emissions studies in SOCMI and the petroleum refining industry.
Fugitive emission sources in SOCMI and in the petroleum refining
industry are similar. Considerable data exist concerning both the
incidence and magnitude of fugitive emissions from petroleum refineries.
Studies of fugitive emissions in SOCMI have been undertaken by EPA to
support the use of emission factors generated during studies of emissions
in petroleum refineries for similar sources in the Synthetic Organic
Chemicals Manufacturing Industry. The results of the EPA SOCMI studies,
EPA data from a study of fugitive emissions from petroleum refineries,
and some industry studies of fugitive emissions are discussed in Section C.I.
Section C.2 consists of the results of three studies on the effects
of maintenance on reducing fugitive VOC emissions from valves in petroleum
refineries and one study on maintenance of valves in a SOCMI process
unit. These results are included as an indication of the reduction in
emissions which could be expected as a function of the designated action
level, and by applying routine on-line maintenance procedures.
C.I FUGITIVE EMISSIONS TEST PROGRAMS
Three SOCMI test programs have been conducted by EPA. One was a
study performed by Monsanto Research Corporation of a small number of
fugitive emission sources in four SOCMI units. More intensive screening
was performed at six SOCMI units in another study. The third EPA study
of SOCMI fugitive emissions was a screening and sampling program conducted
C-l
-------
at twenty-four SOCMI units. The results of these studies are presented
in this section. Similar types of studies have been performed by industry.
This section also contains the results of an Exxon study of fugitive
emissions in cyclohexane unit and a DuPont study of fugitive emissions
in unidentified process units.
The results of a study on fugitive emissions from petroleum refineries
are also presented in this section. Data on fugitive emissions were
obtained from 64 units in thirteen refineries located in major refining
areas throughout the country. Data on the effects of maintenance were
obtained at the last four of these refineries. These results are presented
later in Section C.2 of this Appendix.
C.I.I Study of Fugitive Emissions At Four SOCMI Units1
Monsanto Research Corporation conducted an EPA-IERL sponsored .study
of fugitive emissions at four SOCMI units. The process units were
monochlorobenzene, butadiene, ethylene oxide/glycol, and dimethyl
terephthalate. Due to the small number of plants/processes sampled and
the experimental design of this study, the results were not considered
to be comparable with the results of other studies. Since the data
generated by the MRC study could not be considered representative of
the SOCMI and valid conclusions could not be drawn concerning the
relative magnitude of fugitive emissions in the SOCMI, the results of
the study were not used in the development of standards for fugitive
emissions control. This study demonstrated the need for more intensive
sampling and screening which was undertaken by EPA.
C.I.2 Description and Results of EPA Study of Six SOCMI Units2'3'4'5
The objective of this test program was to gather data on the percen-
tage of sources which leak (as defined by a VOC concentration at the
leak interface of >10,000 ppmv calibrated with methane). To achieve
this objective, an attempt was made to screen all potential leak sources
(generally excluding flanges) on an individual component basis with a
portable organic vapor analyzer. The test crews relied on plant personnel
to identify equipment handling organics. Normally, all pumps and compressor
seals were examined, and the percentage of valves carrying VOC which were
C-2
-------
screened ranged from 33 to 85 percent. All tests were performed with a
Century Systems Corporation Organic Vapor Analyzer, Model 108, with the
probe placed as close to the source as possible. The results of this
study are shown in Table C-l.
Six chemical process units were screened. Unit A is a chlorinated
methanes production facility in the Gulf Coast area which uses methanol
as feedstock material. The individual component testing was conducted
during September 1978. Unit B is a relatively small ethylene production
facility on the West Coast which uses an ethane/propane feedstock.
Testing was conducted during October 1978. Unit C is a chlorinated
methanes production facility in the Midwest. This plant also uses
methanol as the basic organic feedstock. Over the last few years,
several pieces of equipment have been replaced with equipment the company
feels is more reliable. In particular, the company has installed certain
types of valves which they have found do not leak "as much" as other
valves. The individual component testing was conducted during January
1979. Unit D is an ethylene production facility on the Gulf Coast,
using an ethane/propane feed. The facility is associated with a major
refinery, and testing was conducted during March 1979. Units E and F
are part of an intermediate size integrated petroleum refinery located
in the North Central United States. Testing was conducted during November
1978. Unit E is an aromatics extraction unit that produces benzene,
toluene, and xylene by extraction from refined petroleum feedstocks.
Unit E is a new unit and special attention was paid during the design
and startup to minimize equipment leaks. All valves were repacked
before startup (adding 2 to 3 times the original packing) and all pumps
in benzene service had double mechanical seals with a barrier fluid.
Unit F produces benzene by hydrodealkylation of toluene. Unit F was
originally designed to produce a different chemical and was redesigned
to produce benzene.
C-3
-------
TABLE C-l. FREQUENCY OF LEAKS FROM FUGITIVE EMISSION SOURCES IN
SYNTHETIC ORGANIC CHEMICAL UNITS (Six Unit Study)
.Unit Ac
Chloromethanes
Equipment type
Valves
Open-ended 1 ines
Pump seals
i
-P» Compressor seals
Control valves
Pressure relief valves
Flanges
Drains
Number
of
sources
tested
600
52
47
_a
52
7
30
_a
Percent with
screening
values
>1 0,000 ppmv
1
2
15
6
0
3
- - Unit Bc
Ethyl ene
Number Percent with
of screening
sources values
tested >10,000 ppmv
2301 19
386 11
51 21
42 59
128 20
a
a
a
Unit Cd
Chloromethanes
Number Percent with
of screening
sources values
tested >10,000 ppmv
658 0.1
a
39 3
3 33
25 0
a
a
a
_ Unit De__ . Unit Ef
Ethyl ene BTX Recovery
Number Percent with Number Percent with
of screening of screening
sources values sources values
tested >1 0,000 ppmv tested >! 0,000 ppmv
862
90
63
17
25
a
_a
39
14 715 1.1
13 33 0.0
33 33b 3.0
6
44 53 4.0
_a
_a
10 -a
Unit Ff
Toluene HDA
Number Percent with
of screening
sources values
tested >1 0,000 ppmv
427 7.0
28 11.0
30 10.0
_a
44 11.0
a
_a
a
aNo Data
Pump seals In benzene service have double mechanical seals
""Source: Reference 6
Source: Reference 7
eSource: Reference 8
Source: Reference 9
-------
C.I.3 Description and Results of an EPA Study of 24 SOCMI Units10
The U.S. EPA Industrial Environmental Research Laboratory coordinated
a study to develop information about fugitive emissions in the SOCMI. A
total of 24 chemical process units were selected for this purpose. The
process units were selected to represent a cross section of the population
of the SOCMI. ^Factors considered during process unit selections included
annual production volume, number of producers, volatility, toxicity, and
value of the final products. Table C-2 shows the process unit types
selected for screening.
The screening work began with the definition of the process unit
boundaries. All feed streams, reaction/separation facilities, and
product and by-product delivery lines were identified on process flow
diagrams and in the process unit. Process data, including stream
composition, line temperature, and line pressure, were obtained for all
flow streams. Each process stream to be screened was identified and
process data was obtained with the assistance of plant personnel, in
most cases. Sources were screened by a two-person team (one person
handling the hydrocarbon detector and one person recording data).
The Century Systems Models OVA-108 and OVA-128 hydrocarbon detectors
were used for screening. The HNU Systems, Inc., Model PI 101 Photoionization
Analyzer was also used to screen sources at the formaldehyde process
unit. The detector probe of the instrument was placed directly on those
areas of the sources where leakage would typically occur. For example,
gate valves were screened along the circumference of the annular area
around the valve stem where the stem exits the packing gland and at the
packing gland/valve bonnet interface. All process valves, pump seals,
compressor seals, agitator seals, relief valves, process drains, and
open-ended lines were screened. From five to twenty percent of all
flanges were randomly selected and screened. For the purpose of this
program "flange" referred to any pipe-to-pipe or tubing-to-tubing connection,
excluding welded joints.
C-5
-------
Table C-2. Twenty-four Chemical Process Units Screened for
Fugitive Emissions
Unit Type
1. Vinyl Acetate
2. Ethylene
3. Vinyl Acetate
4. Ethylene
5. Cumene
6. Cumene
7. Ethylene
8. Acetone/Phenol
9. Ethylene Dichloride
10. Vinyl Chloride Monomer
11. Formaldehyde
12. Ethylene Dichloride
13. Vinyl Chioride Monomer
14. Methyl Ethyl Ketone
15. Methyl Ethyl Ketone
16. Acetaldehyde
17. Methyl Methacrylate
18. Adipic Acid
19. Trichl oroethylene/Perchloroethylene
20. 1,1,1-Trichloroethane
21. Ethylene Dichloride
22. Adipic Acid
23. Acrylonitrile
,?4. Acrylonitrile
Source: .Reference 11
C-6
-------
Each screening instrument was calibrated on a daily basis, at a
minimum. The model OVA-108 instruments, with a logarithmic scale
reading from 1 ppmv to 10,000 ppmv, were calibrated with high (8,000
ppmv) and low (500 ppmv) concentration methane-in-air standards to
ensure accurate operation at both ends of the instrument's range. The
model OVA-128 instruments, with a linear readout ranging from 0 ppmv to
1,000 ppmv, were also calibrated with high and low concentration standards,
A pre-calibrated dilution probe was required with the OVA-128 when
calibrating with the 8,000 ppmv standard.
The HNU Photoionization instrument, used to screen the formaldehyde
process unit, was calibrated with isobutylene, which has an ionization
potential close to that of formaldehyde.
Results of the screening program at the 24 process units are
summarized in Table G-3.
The fugitive emission sources in the study were screened at an
average rate of 1.7 minutes per source for a two-person team (or 3.4
person-minutes per source). This average screening rate includes time
spent for instrument calibration and repair. Table C-4 presents screening
time data on a unit-by-unit basis. These time requirements are somewhat
higher than would be expected for routine monitoring because of the
extensive record keeping associated with the screening project.
1 9
C.1.4 Description and Results of Refinery Fugitive Emissions Study
Data concerning the leak frequencies and emission factors for
various fugitive sources were obtained primarily at nine refineries.
More complete information for compressors and relief valves emissions
was obtained by sampling at four additional refineries. Refineries were
selected to provide a range of sizes and ages and all of the major
petroleum refinery processing units were studied. The type of process
units and the number of each studied in the first nine refineries are
listed in Table C-5.
C-7
-------
TABLE C-3. SUMMARY OF SOCMI PROCESS UNITS FUGITIVE EMISSIONS
(Twenty-four Unit Study)
o
i
00
Source Type
Flanges
Process Drains
Open Ended Lines
Agitator Seals
Relief Valves
Valves
Pumps
Compressors
Other3
Service
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Gas
Light
Heavy
Light
Heavy
Gas
Gas
Light
Heavy
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
(1)
Number
Screened
1,443
2,897
607
83
527
28
923
3,603
477
7
8
1
85
69
3
9,668
18,294
3,632
647
97
29
19
33
2
(2)
% Not
Screened
4
2
2
23
1
0
17
10
21
46
11
66
72
40
66
17
12
9
4
.6
.6
.4
jl
.9
.0
.5
.4
.5
.1
.1
.7
.7
.5
.7
.5
.2
.9
.3
40.5
9
9
19
33
.4
.5
.5
.3
(3) (4)
% of Screened Sources 95% Confidence Interval
with Screening Values for Percentage of Sources
^10,000 ppmv >10,000 ppmv
4
1
0
2
3
7
5
3
1
14
0
0
3
2
0
11
6
0
8
2
6
21
6
0
.6
.2
.0
.4
.8
.1
.8
.9
.3
.3
.0
.0
.5
.9
.0
.4
.4
.4
.8
.1
.9
.0
.1
.0
(3
(0
(0
(0
(2
(0
(4
(3
(0
(0
(0
(0
(0
o
(0
(10
(6
(0
(6
(0
(0
(6
(0
(0
.6,
.9,
.0,
.3,
.3,
.9.
.4,
.3,
.5,
-4,
.0,
.0.
.7,
.3,
.0,
.8,
.1.
.2,
.6,
.3.
.9,
.0,
.7,
.0,
5.8)
1.8)
0.6)
8.4)
5.8)
23.5)
7.5)
4.6)
2.8)
57.9)
36.9)
97.5)
10.0
10.1
70.8
12.1)
6.8)
0.7)
11.1)
7.3)
22.8)
45.6)
20.2)
84.2)
Includes filters, vacuum breakers, expansion joints, rupture disks, sight glass seals, etc.
nnrre- Bef IT
Source: Ref. 13
-------
TABLE C-4. AVERAGE FUGITIVE EMISSION SOURCE SCREENING RATES
(Twenty-four Unit Study)
Average Screening
Number of Time Per
Process Unit Type Screened Sources Source, Minutes
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Vinyl Acetate
Ethyl ene
Vinyl Acetate
Ethyl ene
Cumene
Cumene
Ethyl ene
Acetone/Phenol
Ethylene Bichloride
Vinyl Chloride Monomer
Formaldehyde
Ethylene Di chloride
Vinyl Chloride Monomer
Methyl Ethyl Ketone
Methyl Ethyl Ketone
Acetal dehyde
Methyl Methacrylate
Adi pic Acid
Tri chol oroethylene/Perchloroethyl ene
1,1, 1-Tri chl oroethane
Ethylene Di chloride
Adipic Acid
Acrylonitrile
Acrylonitrile
1,391
5,078
2,780
5,278
1,025
1,573
3,685
3,207
1,430
868
230
744
2,619
585
679
1,148
2,019
1,577
2,720
570
42
664
1,406
1,864
2.0
1.3
0.9
1.5
0.9
1.0
1.9
3.2
2.6
1.8
1.6
1.6
2.2
1.2
0.9
0.7
1.6
1.9
2.5
1.9
Total 43,182 1.7
aAverage source screening time was determined for a two-person team,
one person screening with a portable hydrocarbon detector and one
person recording data. Average screening time includes time spent
for instrument calibration, maintenance, and repair.
Source: Ref. 14
C-9
-------
TABLE C-5. SAMPLED PROCESS UNITS FROM NINE REFINERIES
DURING REFINERY STUDY
Refinery process unit
Atmospheric distillation
Vacuum distillation
Thermal operations (coking)
Catalytic cracking
Catalytic reforming
Catalytic hydrocracking
Catalytic hydrorefining
Catalytic hydrotreating
Alkylation
Aromatics/isomeri zation
Lube oil manufacture
Asphalt manufacture
Fuel gas/light-ends processing
LPG
Sulfur recovery
Other
Number of
sampled units
7
4
2
5
6
2
2 .
7
6
3
2
1
11
2
1
3
Source: Ref. 15
C-10
-------
In each refinery, sources in six to nine process units were selected
for study. The approximate number of sources selected for study and
testing in each refinery is listed below:
Valves 250-300
Flanges 100-750
Pump seals 100-125
Compressor seals 10-20
Drains 20-40
Relief Valves 20-40
There were normally 500-600 sources selected in each refinery.
The distribution of sources among the process units was determined
before the selection and testing of individual sources was begun.
Individual sources were selected from piping and instrumentation diagrams
or process flow diagrams before a refinery processing area was entered.
Only those preselected sources were screened. In this way, bias based
on observation of individual sources was theoretically eliminated.
The screening of sources was accomplished with portable organic
vapor detectors. The principal device used in this study was the J. W.
Bacharach Instrument Co. "TLV Sniffer" calibrated with Hexane. The
components were tested on an individual basis and only those components
with VOC concentrations in excess of 200 ppmv were considered for further
study.
A substantial portion of these leaking sources were enclosed and
sampled to determine both the methane and nonmethane emission rates. An
important result of this program was the development of a correlation
between the maximum observed screening value (VOC concentration) and the
measured nonmethane leak rate.
Emission factors and leak frequency information generated during
this study are given in Table C-6.
C.I.5 Comparison of Fugitive Emissions Test Data from Refineries and
SOCMI Units
The results of the SOCMI studies and those of the refinery emissions
study are compared in Table C-7.
C-ll
-------
TABLE C-6. LEAK FREQUENCIES AND EMISSION FACTORS FROM FUGITIVE
SOURCES IN PETROLEUM REFINERIES
Equipment
type
Val ves
Gas service
Light 1 iquid service
Heavy liquid service
Pump seals
Light liquid service
Heavy liquid service
Percent of
sources having
screening values
^10,000 ppmv
TLV-Hexane
NA
10
• 12
0
NA
23
2
Estimated emission
factor for
refinery sources,
kq/hr-source
NA
0.021
0.010
0.0003
NA
0.12
0.02
Compressor seals (hydrocarbon 33 0.44
service)
Pressure relief valves 8 0.086
Gas service 0.16
Light liquid service 0.006
Heavy liquid service 0.009
Flanges 0 0.0003
Open-ended lines NA NA
Gas service 0.025
Light liquid service 0.014
Heavy liquid service 0.003
Source: Ref.17
C-12
-------
TABLE C-7. COMPARISON OF LEAK FREQUENCIES FOR FUGITIVE EMISSION
SOURCES IN SOCMI UNITS AND PETROLEUM REFINERIES
Equipment Type
Valves (all)
Gas
Light Liquid
Heavy Liquid
Open-ended 1 ines (all )
Gas
Light Liquid
Heavy Liquid
Pumps (all)
Light Liquid
Heavy Liquid
Percent of SOCMI Sources
Having Screening Values
2-10,000 ppmv, OVA-108,
Metnane (six unit study)
11
10
\
17
Percent of SOCMI Sources
Having Screening Values
alO.OOO ppmv, OVA-108
Methane(24 unit study)"
11.4
6.4
0.4
5.8
3.9
1.3
8.8
2.1
Percent of Petroleum
Refinery Sources Having
^Screening Values
10,000 ppmv, TLV - Hexane1"
10
12
0
N/A
23
2
Compressors (Gas)
Pressure Relief Valves (all
43
0
6.9
33
Gas
Light Liquid
Heavy Liquid
Flanges (all) 3
Gas
Light Liquid
Heavy Liquid
Process Drains (all) N/A
Gas
Light Liquid
Heavy Liquid
Agitator Seals (all ) N/A
Gas
Light Liquid
Heavy Liquid
Other • N/A
3.5
2.9
0.0
4.6
1.2
0.0
2.4
3.8
7.1
14.3
0.0
0.0
N/A
N/A
N/A
0
N/A
N/A
N/A
aSource: Ref. 18, 19, 20, 21
bSource: Ref. 22
cSource: 'tef. 23.
ncludes filters, vacuum breakers, expension joints, rupture disks, sight glass seals, etc.
C-13
-------
C.1.6 Description and Results of the DuPont Study
DuPont conducted a program of fugitive emission measurement from
pumps and valves at two of their plants. The processes of the 5 and 10
year old plants were not revealed. The OVA-108 was used for screening
(leak identification) and for leak rate determination (analysis of
collected leak vapors). The leak rate was determined by taking Tedlar
bags partially filled with air and enclosing the leaking valve. The
hydrocarbon concentration in the bags was recorded as a function of
time. Visual estimates of the initial bag volume were assumed to be ±5
percent. Dupont did not have a dilution probe and, therefore, measurements
above 10,000 ppm were not made. Analysis of the data collected indicates
that no significant difference in leak rates exists between manual and
automatic control valves. Significant trends were observed with changes
in product vapor pressure. It also seemed that full open or closed
valve seat positions resulted in lower leak rates than intermediate
positions. The results of the DuPont study are shown in Table C-8.
?4 25
C.I.7 Description and Results of the Exxon Study '
A fugitive emissions study was conducted by Exxon Chemical Company
at the Cyclohexane unit at their Baytown plant. The total number of
valves, pumps and compressor seals, and safety valves were determined.
For all sources, except valves, all of the fugitive emission sources
were sampled. For valves, a soap solution was used to determine leaking
components. All leaking valves were counted and identified as either
small, medium or large leaks. From the set of valves found to be leaking,
specific valves were selected for sampling so that each class of leaking
valves was in approximately the same proportion as it occured in the
cyclohexane unit.
Heat resistant mylar bags or sheets were taped around the equipment
to be sampled to provide an enclosed volume. Clean metered air from the
filter apparatus was blown into the enclosed volume. The sampling train
was allowed to run until a steady state flow was obtained (usually about
15 minutes). A bomb sample was then taken for laboratory analysis (mass
spectrometry). Table C-9 presents the results of the Exxon study.
C-14
-------
TABLE C-8. FREQUENCY OF LEAKS3 FROM FUGITIVE EMISSION
SOURCES IN TWO DuPONT PLANTS.
Equipment
type
Valves
Gas
Light liquid
Heavy liquid
No. of
leakers
48
35
11
1
No. of
non-leakers
741
120
143
478
Percent
leakers
6.1
23.1
7.1
0.2
Pumps
Light
Heavy
liquid
liquid
1
0
36
6
29
2.7
14.3
0
Leak defined as 10,000 ppm or greater.
Source: Ref. 26
C-15
-------
TABLE C-9. FREQUENCY OF LEAKS3 FROM FUGITIVE EMISSION
SOURCES IN EXXON'S CYCLOHEXANE UNIT
Equipment Total
Source in Unit
Valves
Gas 136
light
liquid 201
Safety
valves 15
Pump .
seals0 8
Compressor
seals5 N/A
Screened and
Samp 1 ed
136
100
15
8
N/A
Percent
Leaking
32
15
87
83
100
Emission
factor(kg/hr)
0.017
0.008
0.064
0.255
0.264
99.8% Confidence
Interval (kg/hr)
0.008 - 0.035
0.003 - 0.007
0.013 - 0.5
0.082 - 0.818
0.068 - 1.045
N/A - Not available
a
b
aLeak defined as 10,000 ppm or greater.
Double mechanical seal pumps and compressors were found to have negligible
leaks.
Source: Reference 27,28
; C-16
-------
C.2 MAINTENANCE TEST PROGRAMS
The results of four studies on the effects of maintenance on fugitive
emissions from valves are discussed in this section. The first two
studies were conducted by refinery personnel at the Union Oil Co. refinery
in Rodeo, California, and the Shell Oil Co. refinery in Martinex, California,
These programs consisted of maintenance on leaking valves containing
fluids with vapor pressures greater than 1.5 Reid Vapor Pressure. The
third study was conducted by EPA. Valves were selected and maintained
at four refineries. The fourth study was conducted by EPA at Unit D
(ethylene unit). The study results and a description of each test
program are given in the following sections.
29
C.2.1 Description and Results of the Union Maintenance Study
The Union valve maintenance study consisted of performing undirected
maintenance on valves selected from 12 different process units. Maintenance
procedures consisted of adjusting the packing gland while the valve was
in service. Undirected maintenance consists of performing valve repairs
without simultaneous measurement of the effect of repair on the VOC
concentration detected. This is in contrast to directed maintenace
where emissions are monitored during the repair procedure. With directed
maintenance, repair procedures are continued until the VOC concentration
detected drops to a specified level or further reduction in the emission
level is not possible. Also, maintenance may be curtailed if increasing
VOC concentrations result.
The Union data was obtained with a Century Systems Corporation
Organic Vapor Analyzer, OVA-108. ATI measurements were taken at a
distance of 1 cm from the seal. Correlations developed by EPA have been
used to convert the data from OVA readings taken at one centimeter to
equivalent TLV readings at the leak interface (TLV-0). This facilitates
comparison of data from different studies and allows the estimation of
omission rates based on screening values-leak rate correlations.
The results of the Union study are given in Table C-10. Two sets
of results are provided; the first includes all reparied valves with
before maintenance screening values greater than or equal to 5,300 ppmv
C-17
-------
TABLE C-10. SUMMARY OF MAINTENANCE STUDY RESULTS FROM THE UNION OIL CO.
REFINERY IN RODEO, CALIFORNIA3
o
i
oo
All valves
with initial
screening values
>5300 ppmvb
Number of repairs attempted
Estimated emissions before maintenance, kg/hrc
Estimated emissions after maintenance, kg/hrc
Number of successful repairs (<5300 ppmv after maintenance)
Number of _.v_a.lves with decreased emissions
Number of valves with increased emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased emissions
Percent of valves with increased emissions
133
9.72
4.69
67
124
9
51.8
50.4
93.2
6.8
All valves
with initial
screening values
<5300 ppmv
21
0.323
0.422
--
13
8
-30.5
--
61.9
38.1
Source: Ref. 33.
The value 5300 ppmv, taken with the OVA-108 at 1 cm., generally corresponds to a value of 10,000 ppmv taken
with a "TLV Sniffer" at 0 cm.
-------
(OVA-108), and the second includes valves with before maintenance screening
values below 5,300 ppmv (OVA-108). A screening value of 5,300 ppmv,
obtained with OVA at 1 cm from the leak interface, is equivalent to a
screening value of 10,000 ppmv measured by a Bacharach Instrument Co.
"TLV Sniffer" directly at the leak interface. The OVA-1 cm readings
have been converted to equivalent TLV-0 cm readings because:
1) EPA correlations which estimate leak rates from screening values
were developed from TLV-0 cm data.
2) Additional maintenance study data exists in the TLV-0 cm format.
3) Method 21 specifies 0 cm screening procedures.
The results of this study indicate that maintenance on valves with
initial screening values above 10,000 ppmv (OVA-108) is much more effective
than maintenance on valves leaking at lower rates. In fact, this study
indicates that emissions from valves are reduced by an average of 51.8
percent for valves initially over 5,300 ppmv while valves with lower
initial screening values experienced an increase of 30.5 percent.
31
C.2.2 Description and Results of the Shell Maintenance Study
The Shell maintenance program consisted of two parts. First, valve
repairs were performed on 171 leaking valves. In the second part of the
program, 162 of these valves were rechecked and additional maintenance
was performed. Maintenance consisted of adjusting the packing gland
while the valve was in service. The second part of the program was
conducted approximately one month after the initial maintenance period.
It was not determined whether the maintenance procedures were directed
or undirected, based on the information reported by Shell.
VOC emissions were measured using the OVA-108 and readings were
obtained one centimeter from the source. This data has been transformed
to TLV-0 cm values as was the Union data. And, the same methods of data
analysis described in Section C.2.1 have been applied to the Shell data.
The results of the Shell maintenance study are given in Table C-ll.
C-19
-------
TABLE C.-ll. SUMMARY OF MAINTENANCE STUDY RESULTS FROM THE SHELL OIL COMPANY
REFINERY IN MARTINEZ, CALIFORNIA
March maintenance
Number of repairs attempted
Estimated emissions before maintenance, kg/hrc
Estimated emissions after maintenance, kg/hrc
Number of successful repairs (<5300 ppmv after
maintenance)
Number of valves with decreased emissions
Number of valves with increased emissions
*p Percent reduction in emissions
ro
o
Percent successful repairs
Percent of valves with decreased emissions
Percent of valves with increased emissions
All repaired valves
with initial screening
values ^5300 ppmv"
161
1 1 . 08
2.66
105
161
0
76.0
65.2
100.0
0.0
All repaired valves
with initial screening
values <5300 ppmv
11
0.159
0.0
—
11
0
100.0
--
100.0
0.0
April maintenance
All repaired valves with
initial (March) screening
values >5300 ppmv
152d
2.95
0.421
45
151
1
85.7
83.3
99.3
0.7
All repaired valves with
initial (March) screening
5300 (note nine valves from initial data set not rechecked in April).
elnitial value of 10 of these valves was <1500 ppm-TLV at O.cm.
-------
C.2.3 Description and Results of the EPA Maintenance Study32
Repair data were collected on valves located in four refineries.
The effects of both directed and undirected maintenance were evaluated.
Maintenance consisted of routine operations, such as tightening the
packing gland or adding grease. Other data, including valve size and
type and the processes' fluid characteristics, were obtained. Screening
data were obtained with the Bacharach Instrument Company. "TLV Sniffer"
and readings were taken as close to the source as possible.
Unlike the Shell and Union studies, emission rates were not based
on the screening value correlations. Rather, each valve was sampled to
determine emission rates before and after maintenance using techniques
developed by EPA during the refinery emission factor study. These
values were used to evaluate emissions reduction.
The results of this study are given in Table C-12. Of interest
here is a comparison of the emissions reduction for directed and undirected
maintenance. The results indicate that directed maintenance is more
effective in reducing emissions than is undirected maintenance, particularly
for valves with lower initial leak rates. The results showed an increase
in total emissions of 32.6% for valves with initial screening values
less than 10,000 ppmv which were subjected to undirected maintenance.
However, this increase is due to a large increase in the emission rate
of only one valve.
C.2.4 Description and Results of Unit D (Ethylene Unit) Maintenance Study35
Maintenance was performed by Unit D personnel. VOC concentration
measurements were made using the OVA-108, and readings were obtained at
the closest distance possible to the source. The results of this study
are shown in Table C-13. Directed and undirected maintenance procedures
were used. The results show that directed maintenance results in more
repairs being successfully completed than when undirected maintenance is
used.
C.2.5 Comparison of Maintenance Study Results
Generally speaking, the results of these maintenance programs would
tend to support the following conclusions:
C-21
-------
TABLE C-12. SUMMARY OF EPA REFINERY MAINTENANCE STUDY RESULTS
o
I
ro
Repaired values with initial Repaired values with initial
screening values >10,000 ppmv screening values <10,000 ppmv
Number of valves repaired
Measured emissions before maintenance
kg/hr
Measured emissions after maintenance
. kg/hr
Number of successful repairs
(<10,000 ppmv after maintenance;
Number of valves with decreased
emissions
Number of valves with increased
emissions
Percent reduction in emissions
Percent successful repairs
Percent of valves with decreased
emissions
Percent of valves with increased
emissions
Directed
Maintenance
9
0.107
0.0139
8
9
0
87.0
88.9
100.0
0.0
Undirected Directed Undirected
Maintenance Maintenance Maintenance
23 10 16
1.809 0.0332 0.120
0.318 0.0049 0.159
.13
21 6 15
2 4 1
82.4 85.2 . -32.6
56.5
91.3 60.0 93.8
8.7 40.0 6.3
Source: Ref.36
-------
TABLE C-13. MAINTENANCE EFFECTIVENESS
UNIT D ETHYLENE UNIT BLOCK VALVES
1. Total number of valves with VOC XLO.OOO ppm
from unit survey 121
2. Total number of valves tested for
maintenance effectiveness 46
% Tested 38%
UNDIRECTED MAINTENANCE
3. Total number subjected to repair attempts 37
4. Successful repairs (VOC -10,000 ppm) . 22
/ % Repaired ' 59%
Followup
DIRECTED MAINTENANCE
5. Number of valves unrepaired by undirected 14
maintenance subjected to directed maintenance
6. Number repaired by followup directed maintenance 5
% of unsuccessful repaired by
directed maintenance . 36%
7. Total number repaired based on undirected 27
maintenance subset (3) above
% Repaired 73%
8. Total number of repairs including leaks not 29
found before initial maintenance
Total % repaired 63%
Total % not repaired 37%
Source: Reference 37
C-23
-------
• A reduction in emissions may be obtained by performing
maintenance on valves with screening values above
10,000 ppmv (measured at the source).
• The reduction in emissions due to maintenance of valves
with screening values below 10,000 ppmv is not as
dramatic and may result in increased emissions.
• Directed maintenance is preferable to undirected maintenance
for valve repair.
The information presented in Tables C-10, C-ll, C-12, and C-13 has
been compiled with the objective of placing the data on as consistent a
basis as possible. However, some differences were unavoidable and
others may have gone unrecognized, due to the limited amount of information
concerning the details of methods used in each study. Therefore, care
should be exercised before attempting to draw specific quantitative
conclusions based on direct comparison of the results of these studies.
C-24
-------
C.3 REFERENCES
1. Memo from Tichenor, B. A., EPA:CPB, to Hustvedt, K. C., EPA:CPB.
October 27, 1980. 22 p. SOCMI fugitive emission sampling by
Monsanto.
2. Memo from Muller, C., EPA:CPB, to file. January 18, 1979. 5 p.
Fugitive emissions data from Plant A and Plant B.
3. Memo from Muller, C. K., EPArCPB, to file. March 19, 1979. 3 p.
Fugitive emissions data from Plant C.
4. Memo from Muller, C., EPA:CPB, to file. March 19, 1979. 11 p.
Fugitive emissions data from Plant D.
5. Trip report. Hustvedt, K. C., EPA:CPB, to Durham, J. F., EPA:CPB.
January 5, 1979. 2 p. Report of November 13-17, 1978 visit to
Plant E and Plant F.
6. Reference 2.
7. Reference 3.
8. Reference 4.
9. Reference 5.
10. Blacksmith, J. R., et al. (Radian Corporation.) Problem Oriented
Report: Frequency of Leak Occurrence for Fittings in Synthetic
Organic Chemical Plant Process Units. (Prepared for U. S. Environmental
Protection Agency.) Research Triangle Park, N. C. EPA Contract
No. 68-02-3171. September 1980.
11. Reference 10.
12. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979.
13. Reference 10.
14. Reference 10.
15. Reference 12.
C-25
-------
16. Meeting Report. Honerkamp, R., Radian Corporation, to Hustvedt,
K. C., EPA:CPB, and distribution list. June 12, 1979. 14 p.
Minutes of meeting between EPA and DuPont representatives about
fugitive emission sampling.
17. Reference 12.
18. Reference 2.
19. Reference 3.
20. Reference 4.
21. Reference 5.
22. Reference 10.
23. Reference 12.
24. Letter and attachment from Cox, J. B., Exxon Chemical Company, to
Weber, B., EPA:CPB. February 21, 1978. 4 p. Copy of letter about
cyclohexane unit fugitive loss data to Hydroscience.
25. Letter and attachment from Cox, J. B., Exxon Company, to Walsh,
R. T., EPAiCPB. March 21, 1979. 4 p. Information about cyclohexane
unit.
26. Reference 16.
27. Reference 24.
28. Reference 25.
29. Letter and attachments from Bottomley, F. R., Union Oil Company, to
Feldstein, M., Bay Area Air Quality Management District. April 10,
1979. 36 p. Information about valve repairability.
30. Reference 12.
31. Letter and attachments from Thompson, R. M., Shell Oil Company, to
Feldstein, M., Bay Area Air Quality Management District. April 26,
1979. 46 p. Information about valve repairability.
32. Radian Corporation. Assessment of Atmospheric Emissions from
Petroleum Refining, Appendix B: Detailed Results. (Prepared for
U. S. Environmental Protection Agency.) Research Triangle Park, N.
C. Publication No. EPA-600/2-80-075c. April 1980.
33. Reference 29.
C-26
-------
34. Reference 31.
35. U. S. Environmental Protection Agency. Air Pollution Emission Test
at Phillips Petroleum Company. Research Triangle Park, N. C. EMB
Report No. 78-OCM-12E. December 1979.
36. Reference 32.
37. Reference 35.
C-27
-------
-------
APPENDIX D - EMISSION MEASUREMENT AND CONTINUOUS MONITORING
D.I EMISSION MEASUREMENT METHODS
To develop data in support of standards for the control of fugitive
emissions, EPA conducted leak surveys at six petroleum refineries and
three synthetic organic chemical manufacturing plants. The resulting
leak determination procedures contained in Reference Method 21 were
developed during the course of this test program.
Prior to the first test, available methods for measurement of
fugitive leaks were reviewed, with emphasis on methods that would provide
data on emission rates from each source. To measure emission rates,
each individual piece of equipment must be enclosed in a temporary cover
for emission containment. After containment, the leak rate can be
determined using concentration change and flow measurements. This
procedure has been used in several studies,^'^ and has been demonstrated
to be a feasible method for research purposes. It was not selected for
this study because direct measurement of emission rates from leaks is a
time-consuming and expensive procedure requiring about $40 and ? manhours
per source. It is not feasible or practical for routine testing because
of the large number of sources within each process unit. There can be
more than 2000 valves in light liquid and gas service in a process unit.
Procedures that yield qualitative or semi-quantitative indications
of leak rates were then reviewed. There are essentially two alternatives:
leak detection by spraying each component leak source with a soap solution
and observing whether or not bubbles were formed; and, the use of a
portable analyzer to survey for the presence of increased organic compound
concentration in the vicinity of a leak source. Visual, audible, or
olfactory inspections are too subjective to be used as indicators of
leakage in these applications. The use of a portable analyzer was selected
as a basis for the method because it would have been difficult to establish
D-l
-------
a leak definition based on bubble formation rates. Also, the temperature
of the component, physical configuration, and relative movement of parts
often interfere with bubble formation.
Once the basic detection principle was selected, it was then necessary
to define the procedures for use of the portable analyzer. Prior to
performance of the first field test, a procedure was reported that
conducted surveys at a distance of 5 cm from the components. This
information was used to formulate the test plant for initial testing.
In addition, measurements were made at distances of 25 cm and 40 cm on
three perpendicular lines around individual sources. Of the three
distances, the most repeatable indicator of the presence of a leak was a
measurement at 5 cm, with a leak definition concentration of 100 or
1000 ppmv. The localized meteorological conditions affected dispersion
significantly at greater distances. Also it was difficult to define a
leak at greater distances because of the small changes from ambient
concentrations observed. Surveys were conducted at 5 cm from the source
during the next three facility tests.
The procedure was distributed for comment in a draft control techniques
guideline documents. Many commentors felt that a measurement distance
of 5 cm could not be accurately repeated during screening tests. Since
the concentration profile is rapidly changing between 0 and about 10 cm
from the source, a small variance from 5 cm could significantly effect
the concentration measurement. In response to these comments, the
procedures were changed so that measurements were made at the surface of
the interface, or essentially 0 cm. This change required that the leak
definition level be increased. Additional testing at two refineries and
three chemical plants was performed by measuring volatile organic concentrations
at the interface surface.
A complication that this change introduces is that a very small
mass emission rate leak ("pin-hole leak") can be totally captured by the
instrument and a high concentration result will be obtained. This has
occurred occasionally in EPA tests and a solution to this problem has
not been found.
D-2
-------
The calibration basis for the analyzer was evaluated. It was
recognized that there are a number of potential vapor stream components
and compositions that can be expected. Since all analyzer types do not
respond equally to different compounds, it was necessary to establish a
reference calibration material. Based on the expected compounds and the
limited information available on instrument response factors, hexane was
chosen as the reference calibration gas for EPA test programs. At the
5 cm measurement distance, calibrations were conducted at approximately
100 or 10QO ppmv levels. After the measurement distance was changed,
calibrations at 10,000 ppmv levels were required. Comments received
indicated that hexane standards at this concentration were not readily
available commercially. Consequently, modifications were incorporated
to allow alternate standard preparation procedures or alternate calibration
gases in the test method recommended in the Control Techniques Guideline
Document for Petroleum Refinery Fugitive Emissions. Since that time,
-additional studies have begun to develop response factor data for two
instrument types. Based on preliminary results, it appears that methane
is a more representative reference calibration material at 10,000 ppmv
levels. Based on this conclusion, and the fact that methane standards
are readily available at the necessary calibration concentration, the
recommended calibration material for this regulation was changed to
methane. ' '
The alternative of specifying a different calibration material for
each type stream and normalization factors for each instrument type was
not intensively investigated. There are at least four instrument types
available that might be used in this procedure, and there are a large
number of potential stream compositions possible. The amount of prior
knowledge necessary to develop and subsequently use such factors would
make the interpretation of results prohibitively complicated. Based on
EPA test results, the number of concentration measurements in the range
where a variability of two or three would change the decision as to
whether or not a leak exists is small in comparison to the total number
of potential leak sources.
D-3
-------
An alternative approach to leak detection was evaluated by EPA
during field testing. The approach used was an area survey, or walkthrough,
using a portable analyzer. The unit area was surveyed by walking through
the unit positioning the instrument probe within 1 meter of all valves
and pumps. The concentration readings were recorded on a portable strip
chart recorder. After completion of the walkthrough, the local wind
conditions were used with the chart data to locate the approximate
source of any increased ambient concentrations. This procedure was
found to yield mixed results. In some cases, the majority of leaks
located by individual component testing could be located by walkthrough
surveys. In other tests, prevailing dispersion conditions and local
elevated ambient concentrations complicated or prevented the interpretation
of the results. Additionally, it was not possible to develop a general
criteria specifying how much of an ambient increase at a distance of
1 meter is indicative of a 10000 ppm concentration at the leak source.
Because of the potential variability in results from site to site,
routine walkthrough surveys were not selected as a reference or alternate
test procedure.
D.2 CONTINUOUS MONITORING SYSTEMS AND DEVICES
Since the leak determination procedure is not a typical emission
measurement technique, there are no continuous monitoring approaches that
are directly applicable. Continual surveillance is achieved by repeated
monitoring or screening of all affected potential leak sources. A
continuous monitoring system or device could serve as an indicator that a
leak has developed between inspection intervals. EPA performed a limited
evaluation of fixed-point monitoring systems for their effectiveness
in leak detection. The systems consisted of both remote sensing devices
with a central readout and a central analyzer system (gas chromatograph)
with remotely collected samples. The results of these tests indicated
that fixed point systems were not capable of sensing all leaks that were
found by individual component testing. This is to be expected since
these systems are significantly affected by local dispersion conditions
D-4
-------
and would require either many individual point locations, or very low
detection sensitivities in order to achieve sjmilar results to those
obtained using an individual component survey.
It is recommended that fixed-point monitoring systems not be required
since general specifications cannot be formulated to assure equivalent
results, and each installation would have to be evaluated individually.
D.3 .PERFORMANCE TEST METHOD
The recommended fugitive emission detection procedure is Reference
Method 21. This method incorporates the use of a portable analyzer to
detect the presence of volatile organic vapors at the surface of the
interface where direct leakage to atmosphere could occur. The approach
of this technique assumes that if an organic leak exists, there will be
an increased vapor concentration in the vincinity of the leak, and that
the measured concentration is generally proportional to the mass emission
rate of the organic compound.
An additional procedure provided in Reference Method 21 is for the
determination of "no detectable emissions". The portable VOC analyzer
is used to determine the local ambient VOC concentration in the vincinity
of the source to be evaluated, and then a measurement is made at the
surface of the potential leak interface. If a concentration change of
less than 2 percent of the leak definition is observed, then a "no
detectable emissions" condition exists. The definition of 2 percent of
the leak definition was selected based on the readability of a meter
scale graduated in 2 percent increments from 0 to 100 percent of scale,
and not necessarily on the performance of emission sources. "No detect-
able emissions" would exist when the observed concentration change
between local ambient and leak interface surface measurements is less
than 200 ppmv.
Reference Method 21 does not include a specification of the instrument
calibration basis or a definition of a leak in terms of concentration.
Based on the results of EPA field tests and laboratory studies, methane
is recommended as the reference calibration basis for fugitive emission
sources in synthetic organic chemical manufacturing industries.
D-5
-------
There are at least four types of detection principles currently
available in commercial portable instruments. These are flame ionization,
catalytic oxidation, infrared absorption (NDIR), and photoionization.
Two types (flame ionizat'ion and catalytic oxidation) are known to be
available in factory mutual certified versions for use in hazardous
atmospheres.
The recommended test procedure includes a set of design and operating
specifications and evaluation procedures by which an analyzer's performance
can be evaluated. These parameters were selected based on the allowable
tolerances for data collection, and not on EPA evaluations of the performance
of individual instruments. Based on manufacturers' literature specifications
and reported test results, commercially available analyzers can meet
these requirements.
The estimated purchase cost for an analyzer ranges from about
$1,000 to $5,000 depending on the type and optional equipment. The cost
of an annual monitoring program per unit, including semiannual instrument
tests and reporting is estimated to be from $3,000 to $4,500. This
estimate is based on EPA contractor costs experienced during previous
test programs. Performance of monitoring by plant personnel may result
in lower costs. The above estimates do not include any costs associated
with leak repair after detection.
D-6
-------
D.4 REFERENCES
1. Joint District, Federal, and State Project for the Evaluation of
Refinery Emissions. Los Angeles County Air Pollution Control
District, Report Numbers 2, 3, 5, 6, and 8. 1957-58.
2. Wetherold, R. and L. Provost. (Radian Corporation.) Emission
Factors and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. Publication No. EPA-600/2-79-044.
February 1979.
3. Memo from Harris, G. E., Radian Corporation, to Wilkins, G. E.,
Radian Corporation. June 19, 1980. 1 p. Information about bagging
costs.
4. Telecon. Harrison, P., Meteorology Research, Inc., with Hustvedt,
K. C., EPAiCPB. January 6, 1977. 3 p. Conversation about refinery
miscellaneous hydrocarbon source sampling.
5. U. S. Environmental Protection Agency. Air Pollution Emission Test
at Atlantic Richfield Company. Research Triangle Park, N. C. EMB
Report No. 77-CAT-6. December 1979.
6. U. S. Environmental Protection Agency. Control of Volatile Organic
Compound Leaks from Petroleum Refinery Equipment. Research Triangle
Park, N. C. Publication No. EPA-450/2-78-036. June 1978.
7. Letter and attachments from McClure, H. H., Texas Chemical Council,
to Barber, W., EPA:OAQPS. June 30, 1980. Comments on the SOCMI
Background Information Document.
D-7
-------
-------
APPENDIX E. METHODOLOGY FOR ECONOMIC ANALYSIS
-------
-------
APPENDIX E: METHODOLOGY FOR COMPUTING COST OF CAPITAL
TO SYNTHETIC ORGANIC CHEMICAL MANUFACTURERS
This appendix describes the process used to estimate the cost of
capital for the chemical industry. The cost of capital for any new project
is the cost of equity, debt, and preferred stock, weighted by the percent-
age of funds generated by each type of financing; that is,
kc = ke f * ki I + kp I
where
k = cost of capital,
kg = cost of equity capital,
k. = cost of debt capital,
k = cost of preferred stock capital,
E = the amount of equity used to finance a given investment,
D = the amount of debt used to finance a given investment,
P = the amount of preferred stock used to finance a given
investment,
I = the total funds needed for the investment.
The k variables are interest rates representing the aftertax return on
investment that is needed to pay stock dividends and interest on debt.
Each k term is a nominal interest rate in that it contains an implicit
allowance for inflation. However, the cost of capital computed with equa-
tion (1) is treated in the text as the real dollar interest rate that would
prevail in times of economic stability. The nominal rate is used as though
it were a real rate partly to ensure that estimates of the cost and other
E-l
-------
adverse economic effects of investment in air pollution controls will be
biased upward rather than downward, and partly to avoid miscalculations
that could result from using the wrong inflation rate to convert the nominal
rate to a real rate.
The first step in estimating equation (1) is to determine the relevant
weights for the three types of financing. It is assumed that the proportion
of debt, equity, and preferred stock to be used on any new project will be
the same as currently exists in the firm's capital structure. This implies
that the firm is currently using the optimal mix of financing. Figures for
the three types of funds came from the COMPUSTAT tapes, supplied by Standard
& Poor's Corporation, for each firm's fiscal year ending in 1977. Common
equity included the par value of common stock, retained earnings, capital
surplus, self-insurance reserves, and capital premium, while debt included
all obligations due more than a year from the company's balance sheet date.
Preferred stock represented the net number of preferred shares outstanding
at year-end multiplied by the involuntary liquidating value per share.
The next step in calculating equation (1) is to estimate the cost of
equity financing. Two approaches are commonly used: the results derived
from the capital-asset pricing model (CAPM) and the results derived from
the dividend capitalization model (DCM). The CAPM compares the returns
from a firm's stock with those from the stock market as a whole, while the
DCM evaluates the stream of dividends and the discount rate needed to
arrive at the firm's existing share price. The required return on equity
using the CAPM is:
k = i + p (k -i) (2)
e m
where
i = the expected risk-free interest rate,
k -i = the expected excess return on the market, and
p = the firm's beta coefficient.
The beta coefficient is an historical measure of the extent to which a
firm's stock price fluctuates in relation to an index of the stock market
as a whole, p takes on a value of zero for a stock whose price is constant,
a value of one for a stock whose price follows the same path as an index of
the whole stock market, and a value of greater than (less than) one for a
E-2
-------
stock whose price fluctuates more (less) dramatically than does the general
index. The CAPM is thus a modified regression equation in which 0 is the
slope of a straight line relating k and k .
The required return on equity using the DCM is:
D
ke = I
l
+9 (3)
where
D-, = the dividend expected in period 1,
PQ = the share price at the beginning of period 1,
g = the expected rate of dividend growth, assumed to be constant.
The DCM is developed on the assumptions that (1) the price of a stock is
the present value of anticipated dividends, and that (2) these dividends
grow each year by a fixed percentage that is less than the required return
on equity.
Figures for equation (2) were developed in the following manner. The
expected risk-free rate was assumed equal to the yield on a 3-month Treasury
Bill, as reported in the October 1, 1979, Wall Street Journal. The current
yield was 10.46 percent. This corresponds to the yield from a bond with no
possibility of default and offering no chance of a capital loss and is
therefore riskless. The firm's beta coefficients came from the September
24, 1979, Value Line Investment Survey. The expected excess return equalled
2.9646 percent, the 5-year average (July 1974-June 1979) of the monthly
excess returns on the Standard & Poor's 500 Stock Index multiplied by
twelve.
Figures for equation (3) came from two sources. Both share price and
expected yearly dividends came from figures reported in the October 1,
1979, Wall Street Journal. The growth rate was calculated from data con-
tained on the COMPUSTAT tapes. Note that the use of historical data does
not necessarily make the estimated rate of return on capital inconsistent
with the first quarter 1980 cost data used in this study as both short- and
long-term interest rates are currently in a state of flux. Three different
growth rates were examined: the 5-year average growth of total assets, the
5-year average growth of per share earnings, and the 5-year average growth
of dividends.
E-3
-------
A number of theoretical reasons exist for preferring the CAPM approach ,
to the DCM for estimating the required return on equity, but the figures ::
calculated revealed a more practical justification. Using growth estimated
from per share earnings or dividends resulted in a number of firms having
negative required returns with the DCM method. Although using the growth
in assets resulted in only one firm with a negative required return, several
firms had extremely low returns (less than 10 percent). It is unreasonable
to expect that stockholders would demand a return on their stock that is
less than the existing yield on Treasury Bills, yet all three variants of
the DCM method led to this conclusion for a number of firms. On the basis
of these considerations the CAPM calculations were selected as the required
return on equity.
The third step in estimating equation (1) is calculating the cost of
debt financing. This would be a relatively easy estimation if interest
rates did not change over time. Past yields on old issues of bonds would
suffice. Since interest rates have been fluctuating, it was felt that a
more forward-looking rate was required. The method selected was to take
the average yield as given in the September 3, 1979, Moody's Bond Survey
for the firm's bond ratings class as the necessary yield the firm must
offer on long-term debt. The firm's ratings class came from the September
1979 Moody's Bond Record or the 1979 Moody's Industrial Manual. A small
number of firms were not rated by Moody's. One firm was ranked in Standard
and Poor's Bond Guide and this was used to approximate a Moody's bond
class. For other firms, data concerning bank notes, revolving credit, or
term-loan agreements that tied the interest rate on these types of debt to
the current prime rate were obtained from the 1979 Moody's Industrial Manual
or the Standard & Poor's Corporation Record. These data were taken to
measure the necessary yield on long-term debt for such firms. Table E-l
presents the yields by ratings class and the prime rate (as of October 1,
1979) used for the cost of debt funds.
The yield on long-term debt does not represent the aftertax cost of
debt financing since interest charges are tax deductable. To arrive at the
aftertax cost of debt capital, the yield must be multiplied by 1 minus the
marginal tax rate.
E-4
-------
TABLE E-l. YIELDS BY RATING CLASS FOR COST OF DEBT FUNDS, 1979
(Prime rate = 13.50 %)
Rati
ngs class
AAA
AA
A
BAA
BA
B
Yield (percent)
9.25
9.59
9.72
10.38
11.97
12.395
kn- = k(l - t) (4)
where
k = the yield on bonds,
t = the marginal tax rate.
It is assumed that the firms in the sample are profitable so that taxes
must be paid, and that their marginal tax rate is 48 percent.
The last step in estimating equation (1) is to calculate the cost
of preferred stock financing. Unlike debt, preferred stock does not have a
maturity date so that the current- yield should approximate the yield on new
issues. The yield is:
(5)
where
U - stated annual dividend,
P = the price of a share of preferred stock.*
The figures for dividends and share price came from the October 1, 1979,
Wall Street Journal or, if not included in this source, from the January 1,
1979, listing in the Daily Stock Price Record. A number of firms did not
have their preferred stock listed in either source, yet had preferred stock
in their capital structures. All used less than 15 percent preferred
*Note that as preferred stock dividends do not increase over time the
growth factor required in the discounted cash flow model (equation 3) is
omitted here.
E-5
-------
stock, with the majority using less than 5 percent. For these firms the
aftertax yield on preferred stock was set equal to the pretax yield on
long-term debt.
Table E-2 lists the cost of capital for all 100 firms in the sample
and also includes some of the components of equation (1). These firms :
represent the best available sample of the approximately 600 firms in the
industry. However, it is likely that on the average the firms included in
the sample are larger than the firms excluded, as many small firms do not
have to publish detailed financial records. This potential sample bias may
have resulted in a slight underestimate of the industry's cost of capital
as, in general, because they are (usually) able to reduce their transactions
costs of borrowing and to represent a less risky investment because of
product diversification, larger firms are often able to acquire investment
funds more cheaply than smaller firms.
E-6
-------
TABLE E-2. FINANCIAL DATA FOR 100 CHEMICAL FIRMS1 ll
Name
Abbott Labs
Akzona
Alco Standard Corp.
Al lied Chem Corp.
American Cyanamiri
Armco Steel Corp.
Atlantic Richfield
Beatrice Foods
Bendix Corp.
Bethlehem Steel Corp.
Borden Inc.
Borg-Warner Chem.
Brown Co.
CPC International
Inc.
Celanese Corp.
Charter International
Oil
Cities Service Co.
Combustion
Engineering
Continental Oil
Crompton & Knowles
Dart Indust.
Dayco Corp.
De Soto, Inc.
Diamond Shamrock
Corp.
Dow Chemical
Ou Pont De Nemours
Eastern Gas & Fuel
Associates
Essex Chem. Corp.
Exxon Corp.
FHC Corp.
Cost of
capital
12.014
10.276
12.151
10.091
11.083
10.588
9.749
11.232
11.118
10.913
10.484
11.863
9.813
11.638
10.181
9.175
10.395
11.494
10.881
11.298
10.689
8.270
11.499
9. 790
10.060
11.328
11.605
12.502
11.875
10.183
Return
on
equity
14.018
13.276
13.425
13.721
13.425
13.276
13. 128
12.832
13.425
14.018
12.683
13.128
12.387
13. 128
13.128
14.166
12.980
14.314
13.721
13.425
14.166
12.980
13.128
13.721
14.018
13.573
14.018
14.166
13.276
13.573
Return
on ,
debt"
9.590
10.380
15.120
9.720
9.590
9.720
9.590
9.250
9. 720
9.720
9.590
9.720
12.395
9.590
11.970
12.395
9.720
9.720
9.590
14.450
9.720
11.970
13.750
9.720
9.590
9.250
14.180
12.395
9.250
9.720
Return
on
preferred
stock0
--
--
--
--
--
6.461
--
7.429
3.333
--
--
--
--
--
10.084
--
--
--
2.564
--
4.211
6.071
--
--
--
8.654
--
--
--
6.250
Proportion
of
equity
. 77262
.61914
.64134
.58118
.72252
. 66880
. 51602
. 79803
. 72911
.65360
.71317
.82756
. 56680
.81691
.53511
.27557
.67388
. 68700
. 67568
.53329
.63113
.30351
. 72746
.54639
.56176
.72512
. 63681
.78453
. 83450
.59257
Proportion
of
debt
.216575
. 380859
. 259343
.418825
. 277480
. 306858
.362174
. 194329
. 248140
. 346402
.285155
. 145263
. 433202
. 183087
. 396896
.623167
. 326120
. 296229
.321308
.375634
.231645
.666445
.272535
.453615
.438236
.23217?
. 363188
.215465
. 165504
.339730
Proportion
of
preferred
stock
. 010804
.000
.099317
.000
.000
.024337
. 121802
.007644
. 022754
.000
.001677
.027181
.000
.000
.067997
. 101265
.000
.016774
.003009
.091078
.137221
. 030044
.000
.000
.000
.042712
.000
.000
.000
.067701
(continued)
-------
TABLE E-2. (continued)
oo
Name
Ferro Corp.
Firestone Tire &
Rubber
Ford Motor Co.
GAP Corp.
General Electric Co.
General Motors Corp.
General Tire & Rubber
Georgia-Pacific Corp.
Goodrich (B.F.) Co.
Goodyear Tire &
Rubber Co.
Gulf Oil Corp.
Hercules Inc.
Inland Steel
Insilco Corp.
Interlake, Inc.
International
Harvester
Kaiser Steel Corp.
Kraft Inc.
Harathon Oil Co.
Martin Marietta Chen.
Mead Corp.
Merck & Co.
Minnesota Mining &
Manuf.
Mobil Oil Corp.
Monsanto Co.
Morton-Norwich
Products
National Distillers
& Chen.
National Steel Corp.
Northwest Indust.
Cost of
capital
12. 369
10.610
12.069
9.398
12. 130
12. 798
11.440
10.793
10.430
10. 101
11. 745
11.177
10. 092
9.339
11.331
10.534
11.688
10. 774
9.582
11.238
10.000
12.309
12.572
10.868
10. 970
10.726
11.037
9.909
8.015
Return
on
equity
13 276
12.980
13.276
13.573
13.721
13.425
13.276
13.573
13.276
12.980
12.980
13.869
12.980
13.276
13.128
13.573
14.018
12.683
13. 128
13.276
13.869
13.573
13.869
13.128
13.573
13.721
13.128
12.683
13.869
Return
on a
debt3
9.720
9.720
91250
10.380
9.250
9.250
11.970
9.590
10. 380
9.720
9.250
9.720
9.590
11.970
9.720
9.720
14.000
9.250
9.720
9.720
9.720
9.250
9.250
9.250
9.590
9.. 720
9.720
9.590
10. 380
Return
on
preferred
stock0
--
--
--
7.559
—
8.715
--
--
8.864
--
--
—
--
7.752
--
--
--
—
--
—
4.308
--
--
--
5.000
--
9.193
--
2.9412
Proportion
of
equity
. 88968
. 70096
.85743
. 44490
. 82148
.91962
. 73287
.67625
.62957
.63679
. 84880
. 69461
. 62702
. 41885
.77736
. 63297
.63274
. 75752
. 56074
. 75212
. 56423
.85461
.85677
. 72833
. 69690
.65441
. 73310
. 63946
. 32561
Proportion
of
debt
.110317
. 299038
. 142565
. 387035
. 178521
.063516
. 258968
.323751
. 349707
.363210
.151203
. 305394
. 352735
. 475634
. 222640
. 348230
.345717
. 242479
.439257
.247882
.398718
. 143358
. 143235
.271665
. 300335
. 345589
. 251565
. 360538
.617085
Proportion
of
preferred
stock
.000
.000
.000
. 168061
.000
. 0168G2
.008163
.000
.020723
.000
.000
.000
. 020249
.105511
.000
. 018796
.021539
.000
.000
.000
.037048
.001827
.000
.000
.002767
.000
.015334
.000
. 057301
(continued)
-------
TABLE E-2. (continued)
I
IT)
Name
Owens-Corning
Fiberglass
PPG Industries
Penwalt Corp.
Pfizer
Phillips Petroleum Co.
Procter & Gamble Co.
Quaker Oats Co.
Reeves Bros. Inc.
Reichold Chems.
Republic Steel Corp.
Riegel Textile Corp.
Rockwell International
Rohn and Haas Co.
SCM Corp.
Scott Paper Co.
Shakespeare Co.
Sherwin-Williams Co.
Squibb Corp. >
A. E. Staley Mfg. Co.
Stauffer Chemical Co.
Sterling Drug
Sun Chem. Corp.
Sybron Corp.
Tenneco Inc.
Texaco
Texfi Indust.
Textron Inc.
Union Camp Corp.
Union Carbide Corp.
Union Oil, Calif.
Uni royal
U.S. Gypsum
U.S. Steel Corp.
Upjohn Co.
Cost of
capital
11.653
10.596
9.013
11.244
11.670
11.824
10.946
10.629
10.647
11.305
11.201
9.589
10.739
10.835
10.784
11. 229
9.617
11.266
10.428
10.188
12.595
10.427
10.786
9.155
11.230
10.090
10.085
11.359
10.775
10.577
10.514
10.726
10.919
11.052
Return
on
equity
13.425
13.276
13.276
14.018
13.721
13.276
13.573
12.535
13.425
13.425
12.980
12.535
13.721
14.018
13.721
13.276
12.980
14.018
13.573
13.425
13.276
13.573
13.869
12.980
12.980
13.275
13.425
13.276
13.573
13.128
13.425
13.276
13.573
13.573
Return
on a
debt3
9.720
9.590
9.720
9.590
9.250
9.250
9.720
10.380
10.380
9.720
11.970
9.720
9.720
10.380
9.590
14.000
10.380
9.590 .
9.720
9.720
9.590
12.395
9.720
10.380
9.250
16.000
9.720
9.590
9.590
9.590
11.970
9.590
9.590
9.590
Return
on
preferred
stock0
-- '
--
7.529
--
--
--
9.008
--
--
--
--
5.398
--
'
--
•
10.00
—^
--
--
--
--
3.887
--
--
6.222
--
--
--
16.000
5.539
--
--
Proportion
of
equity
. 78828
.67661
.41712
. 69289
.76982
.82842
. 651578
. 732870
.571986
. 746819
. 736598
. 602132
.655939
.630766
.660791
.658505
. 523981
.695345
.629947
.613351
.917816
. 558689
.616191
. 505890
. 785863
. 356904
.577353
.768639
.674170
.663994
.521603
.686341
. 690912
. 706383
Proportion
of
debt
.211721
.323394
. 369200
.307113
.230179
. 171428
. 262094
.267130
.295871
.253181
.263402
. 309032
. 344061
.369234
.333680
. 341495
.422439
. 304655
. 368508
. 386649
. 082184
.441311
.319517
. 442129
.214137
.643096
.252757
.231361
. 325830
. 295934
.423786
.223477
. 309088
. 293617
Proportion
of
preferred
stock
.000
.000
.213675
.000
.000
.000153
. 086328
.000
.132143
.000
.000
. 088836
.000
.000
.005529
.000
.053579
.000
.001544
.000
.000
.000
.064292
.051981
.000
.000
. 169890
.000
.000
.040072
.054611
.090182
.000
.000
(continued)
-------
TABLE E-2. (continued)
Mm
Vulcan Materials Co.
Walter (J1«) Corp.
Westlnghouse Electric
Corp.
Weyerhaeuser Co.
Wheel ing-Pittsburgh
Steel
Whittaker Corp.
Wit Che*. Corp.
Cost of
capital
10.675
9.019
12.596
10.402
11.238
10.070
10.736
Return
on
equity
12.980
13. 721
14.018
14.166
13.869
14.314
13.573
Return
on a
debt
9.720
11.970
9.720
9.590
14.000
11.970
9.720
Return
on
preferred
stock"
4.444
8.837
5.957
12.739
3.313
Proportion
of
equity
.709218
.398726
.838775
.583685
.512893
.457808
.673790
Proportion
of
debt
.290782
.491966
.155115
.357341
.381136
. 517470
.292825
Proportion
of
preferred
stock
.000
. 109308
.006110
.058973
. 105972
.024722
.033385
*The return on debt data represent pretax estimates and are nultipled by 0.52 to obtain the aftertax rates
used in computing the cost of capital.
''flashes Indicate missing data. In these cases the pretax returns on debt were used to compute the cost
of capital.
-------
APPENDIX E REFERENCES
1. COMPUSTAT. New York: Standard & Poor's Corporation, 1978.
2. Daily Stock Price Record. New York: Standard & Poor's Corporation,
1979.
3. Moody's Bond Record. New York: Moody's Investors Service, Inc.,
September 1979.
4. Moody's Bond Survey. New York: Moody's Investors Service,
September 3, 1979.
5. Moody's Industrial Manual. New York: Moody's Investors Service,
Inc., 1979.
6. Scherer, F. M., et al. The Economics of Multi-Plant Operation.
Cambridge, Mass.: Harvard University Press, 1975.
7. Standard & Poo^s Bond Guide. New York: Standard & Poor's Corpora-
tion, September 1979.
8. Standard & Poor's Corporation Records. New York: Standard & Poor's
Corporation, September 1979.
9. Standard & Poor's Statistical Service. New York: Standard & Poor's
Corporation, October 1979.
10. Value Line Investment Survey. New York: Arnold Bernhard & Co.,
Inc., September 24, 1979.
11. The Wall Street Journal. New York: Dow Jones & Company, October 1,
1979.
E-ll
-------
-------
APPENDIX F - SYNTHETIC ORGANIC CHEMICALS MANUFACTURING INDUSTRY
-------
-------
OCPDB No.* Chemical
20 . Acetal
30 Acetaldehyde
40 Acetaldol
50 Acetamide
65 Acetanilide
70 Acetic acid
80 Acetic anhydride
90 Acetone
100 Acetone cyanohydrin
110 Acetonitrile
120 Acetophenone
125 Acetyl chloride
130 Acetylene
140 Acrolein
150 Acrylamide
160 Acrylic acid and esters
170 Acrylonitrile
180 Adi pic acid
185 Adiponitrile
190 Alkyl naphthalenes
200 ATlyl alcohol
210 Allyl chloride
220 Aminobenzoic acid
*The OCPDB Numbers are reference indices assigned to the various chemicals
in the Organic Chemical Producers Data Base developed by EPA.
F-l
-------
230 Aminoethylethanolamine
235 p-aminophenol
240 Amyl acetates
250 Amyl alcohols
260 Amyl amine
270 Amyl chloride
280 Amyl mercaptans
290 Amyl phenol
300 Aniline
310 Aniline hydrochloride
320 Anisidine
330 Anisole
340 Anthranilic acid
350 Anthraquinone
360 Benzaldehyde
370 Benzamide
380 Rcn/.cne
390 Benzenedisulfonic acid
400 Benzcnesulfonic acid
410 Benzil
420 Benzilic acid
430 Benzoic acid
440 Benzoin
450 Benzonitrile
460 Benzophenone
I
430 Benzotrichloride
F-2
-------
OCPDB No. ChcinicaJ _
490 Benzoyl chloride
500 ' Benzyl alcohol
510 Benzyl amine
520 Benzyl benzoate
530 Benzyl chloride
540 Benzyl dichloride
550 Biphenyl
560 Bisphenol A
570 Bromobenzene
580 Bromonaphthalene
590 Butadiene
592 1-butene
600 n-butyl acetate
630 n-butyl acrylate
640 n-butyl alcohol
650 s-butyl alcohol
660 t-butyl alcohol
670 n-butylainine
680 s-butylamine
690 t-butylamine
700 p-tert-butyl benzoic acid
710 1,3-butylene glycol
750 n-bulyraldohyde
760 Butyric acid
770 Butyric anhydride
780 Butyronitrile
F-3
-------
.QQPDB. ICL . Cjicjiin c_al
785 Caprolactam
790 Carbon disulfide
800 Carbon tetrabromide
810 Carbon tetrachloride
820 Cellulose acetate
840 Chloroacetic acid
850 m-chloroani1ine
860 o-chloroaniline
870 p-chloroaniline
880 Chlorobenzaldehyde
890 Chlorobenzene
900 Chlorobenzoic acid
905 Chlorobenzotrichloride
910 Chlorobenzoyl chlpride
920 Chlorodifluoroethane
921 Chlorodifluoromethane
930 Chloroform
940 Chloronapthalcne
950 o-chlorom'trobenzene
951 p-chloronitrobenzene
960 Chlorophcnols
964 Chloroprene
965 ChlorosulTonic acid
970 m-chlorotoluene
980 o-chlorotoluone
990 p-chlorotoluGne
F-4
-------
OCPDB No. .-.
992 Chlorolrifluoroinethane
1000 m-cresol
1010 o-cresol
1020 p-cresol
1021 Mixed cresols
1030 Cresylic acid
1040 Crotonaldehyde
1050 Crotonic acid
1060 Curncne .
1070 Cumene hydroperoxide
1080 Cyanoacetic acid
1090 Cyanogen .chloride
1100 Cyanuric acid
1110 . Cyanuric chloride
1120 Cyclohexane
1130 Cyclohexanol
1140 Cyclohexanone
1150 Cyclohexene
1160 Cyclohexylamine
1170 Cyclooctadiene
1180 Decanol
1190 Diacetone alcohol
1200 Dianiinobenxoic acid
1210 Dichlorbariiline
1215 m-dichlorobcnzcne
1216 o--dichlorobenzene
F-5
-------
OCPDB No. Chemical
1220 p-dichlorobenzene
1221 Dichlorodifluoromethane
1244 1,2-dichloroethane (EDC)
1240 Dichloroethyl ether
1250 Dichlorohydrin
1270 Dichloropropene
1280 Dicyclohexylanrine
1290 Diethyl amine
1300 Diethylene glycol
1304 Diethylene glycol diethyl ether
1305 Diethylene glycol dimethyl ether
1310 Diethylene glycol monobutyl ether
1320 Diethylene glycol monobutyl ether acetate
1330 Diethylene glycol monoethyl ether
1340 Diethylene glycol monoethyl ether acetate
1360 Diethylene glycol monomethyl ether
1420 Diethyl' sulfate
1430 Difluoroethane
1440 Diisobutylene
1442 Diisodecyl phthalate
1/144 Diisooctyl phthalate
14130 Diketene
1460 Diiiifithyltiiiiine
i
1470 N.N-dimethylaniline
1480 N.N-dimethyl ether
1490 N.N-dimelhylfoniitiinide
F-6
-------
OCPDB No. Chemical
1495 Dimethylhydrazine
1500 Dimethyl sulfate
1510 Dimethyl sulfide
1520 Dimethyl sulfoxide
1530 Dimethyl terephthalate
1540 3,5-dinitrobenzoic acid
1545 ; Dinitrophcnol
.1550 Dinitrotoluene
1560 Dioxane
1570 Dioxolane
1580 Diphenylamine
1590 Diphenyl oxide
1600 •Diphenyl thiourea
1610 Dipropylene glycol
1620 Dodecene
1630 Dodecylaniline
1640 Doclccyl phenol
1650 Epichlorohydrin
1660 L'thanc-1
1661 Etha'nol amines
1670 Ethyl acetate
1680 Ethyl acetoacotate
1690 Ethyl acrylate
1700 Ethyl amine
1710 Ethylbmizone
F-7
-------
OCPDB No. Chemicals
1720 Ethyl bromide
1730 Ethyl cellulose
1740 Ethyl chloride
1750 Ethyl chloroacetate
1760 Ethylcyanoacetate
1770 Ethylene
1780 Ethylene carbonate
1790 ' Ethylene chlorohydrin
1800 Ethylenediamine
1810 Ethylene dibromide
1830 Ethylene glycol
1840 Ethylene glycol diacetate
1870 Ethylene glycol dimethyl ether
1890 Ethylene glycol monobutyl ether
1900 Ethylene glycol monobutyl ether acetate
1910 Ethylene glycol monoethyl ether
1920 EthylCIIR glycol monoethyl ether acetate
1930 Ethylene glycol monomethyl ether
1940 Ethylene glycol monomethyl ether acetate
I960 Ethylene glycol monophcnyl ether
1970 Ethylene glycol monopropyl ether
1980 Ethylene oxide
1990 Ethyl ether
2000 2-ethyihoxanol
2010 Ethyl orthoformate
2020 Ethyl oxalate
F-8
-------
OCPD3__Np_._ Chui mra 1
20:50 Ethyl sodium oxalacctate
2040 Formaldehyde
2050 Fonnamide
2060 Formic acid
2070 Fumaric acid
2073 Furfural
2090 Glycerol (Synthetic)
2091 Glycerol dichlorohydrin
2100 Glycerol triether
2110 Glycine
2120 Glyoxal
2145 Hexachlorobenzene
2150 Hcxachloroethane
2160 Hexndecyl alcohol
2165 Hexamethylenediann'ne
2170 Hcxamethylene glycol
2180 llexiimethylenetctramine
2190 Hydrogen cyanide
2200 Hydroquinane
2210 p-hydroxyhenzoic acid
2240 Isoamylene
2250 Isobutanol
22GO TMjbtiLyl .icctoi.e
2261 Isobutylone
2270 Isobutyraldchyde
2280 Isobutyric acid
F-9
-------
OCPDB Np^ Chemical
2300 Isodccanol
2320 Isooctyl alcohol
2321 Isopentane
2330 Isophorone
2340 Isophthalic acid
2350 Isoprene
2360 Isopropanol
2370 Isopropyl acetate
2380 Isopropylamine
2390 Isopropyl chloride
2400 Isopropylphenol
2410 Ketene
2414 Linear alkyl sulfonate
2417 Linear alkylbenzene
2420 Maleic acid
2430 Maleic anhydride
2440 Malic acid
2450 Mesityl oxide
2455 Metanilic acid
2460 Methacrylic acid
2/190 Methallyl chloride
Z'jQO Metlianol
ZblO Methyl acetate
2520 Methyl aceloacotate
2530 Methylarnine
2540 n-methylaniline
F-10
-------
OCPDB No. Chemi_cal
2545 Methyl bromide
2550 Methyl butynol
2560 Methyl chloride
2570 Methyl cyclohexane
2590 Methyl cyclohexanone
2620 Methylene chloride
2530 Methylene dianiline
2635 Methylene diphony1 diisocyanate
2640 Methyl ethyl ketone
2645 Methyl formate
2650 Methyl isobutyl carbinol
2660 Methyl isobutyl ketone
2665 Methyl methacrylate
2670 Methyl pentynol
2690 a-methylstyrene
2700 Morpholine
2/10 u-n-iph thai one sulfonic acid
2720 p-nanhthalcne sulfonic acid
2730 a-naphthpl
2740 B-naphthol
2750 Neopentanoic acid
2756 o-nitroaniline
2757 p-nitroaniline
2760 o-niLroanisole
2762 p-nitroanisole
2770 Nitrobenzene
F-11
-------
2780 Nitrobenzoic acid (o, n, and p)
2790 Nitroethane
2791 Nitromethane
2792 Nitrophenol
2795 Nitropropane
2800 Nitrotoluene
2810 Nonene •
2820 Nonyl phenol
2830 Octyl phenol
2840 Paraldehyde
2850 Pentaerythritol
2851 n-pentane
2855 1-pentene
2860 Perchloroethylene
2882 Perch!oromethyl mercaptan
2890 o-phenetidine
2900 • p-phenetidine
2910 Phenol
2920 Phenolsulfom'c acids
2930 Phenyl anthranilic acid
2940 Phenylenediamine
2950 Phosgene
2960 Phthalic anhydride
2970 Phthalimide
2973 n-picoline
2976 Piperazine
F-12
-------
OCPDB_ Np^ ____
3000 Polybutcnes
3010 Polyethylene glycol
3025 ' Polypropylene glycol
3063 Propionaldehyde
3066 Propionic acid
3070 n-propyl alcohol
3075 Propylarnine
3080 Propyl chloride
3090 Propylene
3100 Propylene chlorohydrin
3110 Propylene dichloride
3111 Propylene glycol
3120 Propylene oxide
3130 Pyridine
3140 Quinone
3150 Resorcinol.
3160 Resorcylic acid
3170 . Salicylic acid
3180 Sodium acetate
3181 Sodium benzoate
3190 Sodium carboxymethyl cellulose
3191 Sodium chloroacetate
I
3200 Sodium formate
3210 Sodium phunate
3220 Sorbic acid
3230 Styrone
3240 Succinic acid
I
F-13!
-------
OCPDB No. Chemical
3250 Succinitrile
3251 Sulfanilic acid
3260 Sulfolane
3270 Tannic acid
3280 Terephthalic acid
3290 & 3291 Tetrachloroethanes
3300 Tetrachlorophthalic anhydride
3310 Tetraethyllead
3320 Tetrahydronapthalene
3330 Tetrahydrophthalic anhydride
3335 Tetramethyllead
3340 Tetramethylenediamine
3341 Tetramethylethylenediamine
3349 Toluene
3350 Toluene-2,4-diamine
3354 Toluene-2,4-diisocyanate
3355 Toluene diisocyanates (mixture)
3360 Toluene sulfonamide
3370 Toluene sulfonic acids
3380 Toluene sulfonyl chloride
3381 Toluidines
3390,3391 Trichlorobi n/oiH'S
ft 3J93
3395 1,1 ,1-tricliloroi'thane
3-100 1,1,2-trichloroc'th,-ine
F-T4
-------
C:PDB No. Chemical
3410 ' Trichloroethylene
3411 Tn'chlorofluoromethane
3420 1,2,3-trichloropropane
3430 1,1,2-trichloro-l,2,2-trifluoroethane
3450 Triethylamine
3460 Tn'ethylene glycol
3470 Tn'ethylene glycol dimethyl ether
3480 Triisobutylcne
3490 Tririethylamine
3500 Urea
3510 Vinyl acetate
3520 Vinyl chloride
3530 Vinylidene chloride
3540 Vinyl toluene
3541 Xylenes (mixed)
3560 o-xylene
3570 p-xylene
3580 Xylenol
3590 Xylidine
F-15
-------
-------
APPENDIX G - UNCONTROLLED EMISSIONS ESTIMATES
-------
-------
TABLE G-l. UNCONTROLLED EMISSIONS ESTIMATES FROM THE MODEL UNITS'
Fugitive Emission Source
Pumps
Light liquid
Heavy liquid
In-line valves
Vapor service
Light liquid service
Heavy liquid service
Safety/relief valves
Vapor service
Light liquid service
Heavy liquid service
Open-ended valves and lines
Vapor service
Light liquid service
Heavy liquid service
Compressors
Sampling Connections
Flanges
Total from all Fugitive Emission Sources
Model Unit A
kg/hr % of Total
1.1 14
0.96
0.14
2.76 36
1.89
0.84
0.025
1.78 23
1.76
0.006
0.009
1.03 13
0.225
0.66
0.14
0.44 6
0.39 5
0.18 2
7.68
Uncontrolled Emissions
Model Unit B
kg/hr % of Total
4.08 14
3.48
0.60
11.11 38
7.66
3.35
0.10
6.78 23
6.72
0.024
0.036
4.14 14
0.925
2.65
0.57
0.88 3
1.56 5
0.72 2
29.3
Model
kg/hr
12.78
10.92
1.86
34.14
23.46
10.37
0.31
21.0
20.8
0.078
0.13
12.75
2.88
8.13
1.74
3.52
4.80
2.22
91.2
Unit C
% of Total
14
37
23
14
4
5
2
Calculated from the emission factors 1n Table 3-1 and the fugitive emission source counts 1n Table 6-1.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing]
1. REPORT NO.
EPA-45Q/3-80-033a
4iT\J6cE*iN?j5y?TIJL^
2.
3. RECIPIENT'S ACCESSION NO.
SND UBL . .
Fugitive Emissions in Synthetic Organic Chemicals
Manufacturing Industry - Background Information for
Proposed Standards
5. REPORT PATE - • •
November 1980 ;:_._.__ __„
6. PERFORMING ORGANIZATION CODE
7 AUTHORISI
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.
68-02-3058
12. .SPONSORING .AGEJSICY..MAME AMD ADDRESS , r. , ,
DAA for Air Tjuality 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
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Standards of performance to control fugitive emissions of VOC from new, modified,
and reconstructed Synthetic Organic Chemical Manufacturing Industry (SOCMI) plants
are beinn proposed under Section 11.1 of the Clean Air Act. This document contains
information on SOCMI, emission control technology for fugitive emissions of VOC,
Regulatory Alternatives which were considered, analyses of environmental, energy,
costs, and other technical data to support the standard of performance.
17.
II.
KEY1 WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air pollution
Pollution Control
Standards of performance
Volatile Organic Compounds
Organic Chemical Industry
b.IDENTIFIERS/OPEN ENDEDTERMS
Air pollution control
COSATI 1 icld/Ciroup
18. DISTRIBUTION STATEMENT . ., ., , .--.,,
Release unlimited. Available from EPA
Library (MD-35), Research Triangle Park,
North Carolina 27711
19. SECURITY CLASS (This Report/
unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispagfi
unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION IS OBSOLETE
------- |