United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
November 1992
EPA-453/D-92-016c
Air
Hazardous Air Pollutant Draft
Emissions from Process Units EIS
in the
Synthetic Organic Chemical
Manufacturing Industry--
Background Information
for Proposed Standards
Volume 1C: Model Emission Sources
-------
-------
EPA-453/D-92-016c
Hazardous Air Pollutant Emissions
from Process Units in the
Synthetic Organic Chemical
Manufacturing Industry--
Background Information
for Proposed Standards
Volume 1C: Model Emission Sources
Emission Standards Division
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1992
-------
(DISCLAIMER)
This Report has been reviewed by the Emission Standards Division of
the Office of Air Quality Planning and Standards, EPA, and approved
for publication. Mention of trade names or commercial products is
not intended to constitute endorsement or recommendation for use.
Copies of this report are available through the Library Services
Office (MD-35), U.S. Environmental Protection Agency, Research
Triangle Park, N.C. 27711, or from the National Technical
Information Service, 5285 Port Royal Road, Springfield, VA 22161.
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information for Proposed Standards
Hazardous Air Pollutant Emissions from Process Units
in the Synthetic Organic Chemical Manufacturing Industry
Volume 1C: Model Emission Sources
Prepared by:
L
(Date)
Bruce Jordan /~~
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
1. The proposed standards would regulate emissions of organic
hazardous air pollutants (HAP's) emitted from chemical
manufacturing processes of the Synthetic Organic Chemical
Manufacturing Industry (SOCMI). Only those chemical
manufacturing processes that are part of major sources under
Section 112 (d) of the CAA would be regulated. The recommended
standards would reduce emissions of 149 of the organic
- chemicals identified in the CAA list of 189 hazardous air
pollutants.
2. Copies of this document have been sent to the following
Federal Departments: Labor, Health and Human Services,
Defense, Office of Management and Budget, Transportation
Agriculture, Commerce, Interior, and Energy; the National
Science Foundation; and the Council on Environmental Quality
Copies have also been sent to 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 this document is 90 days from the date
of publication of the proposed standard in the Federal
Register. Ms. Julia Stevens may be contacted at 919-541-5578
regarding the date of the comment period.
4. For additional information contact:
Dr. Janet Meyer
Standards Development Branch (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Telephone: 919-541-5299
5. Copies of this document may be obtained from:
U.S. EPA Library (MD-35)
Research Triangle Park, N.C.
27711
-------
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
(NTIS)
-------
OVERVIEW
Emission standards under Section 112(d) of the Clean Air Act
apply to new and existing sources in each listed category of
hazardous air pollutant emission sources. This background
information document (BID) provides technical information used in
the development of the Hazardous Organic National Emission
Standard for Hazardous Air Pollutants (NESHAP), which will affect
the Synthetic Organic Chemical Manufacturing Industry (SOCMI).
The BID consists of three volumes: Volume 1A, National Impacts'
Assessment (EPA-453/D-92-016a); Volume IB, Control Technologies
(EPA-453/D-92-016b); and Volume 1C, Model Emission Sources
(EPA-453/D-92-016C).
Volume 1A presents a description of the affected industry
and the five kinds of emission points included in the impacts
analysis: process vents, transfer loading operations, equipment
leaks, storage tanks, and wastewater collection and treatment
operations. Volume 1A also describes the methodology for
estimating nationwide emissions, emission reductions, control
costs, other environmental impacts, and increases in energy
usage resulting from a potential NESHAP; and presents three
illustrative sets of potential national impacts and a summary of
the economic analysis. While Volume 1A provides the overview of
how information on model emission sources and control technology
cost were used to estimate national impacts, Volumes IB and 1C
contain detailed information on the estimation of control
technology performance and costs and model emission source
development.
Volume IB discusses the applicability, performance, and
costs of combustion devices; collection systems and recovery
devices; storage tank improvements; and control techniques for
equipment leak emissions. These control technologies were the
basis of the Hazardous Organic NESHAP impacts analysis. These
control technologies are applicable to emission points in the
SOCMI and in other source categories. Methods for estimating
-------
capital costs and annual costs (including operation and
maintenance costs) of each control technology are presented.
Volume 1C presents descriptions of each kind of emission
point included in the impacts analysis and the development of
model emission sources to represent each kind of emission point
for use in the impacts analysis. -The emission reductions, other
environmental impacts, and energy impacts associated with
application of the control technologies described in Volume IB to
the model emission sources is discussed. For illustrative
purposes, the environmental, energy, and cost impacts that would
results from control of several example model emission sources
are presented.
-------
TABLE OF CONTENTS
Section
Page
LIST OF TABLES xi
LIST OF FIGURES xiv
ACRONYM AND ABBREVIATION LIST ..... XV
1.0 INTRODUCTION 1-1
2.0 PROCESS VENTS . 2-1
2.1 Emission Source Description . 2-1
2.1.1 Reactor Processes 2-2
2.1.2 Distillation Operations 2-4
2.1.3 Air Oxidation Processes 2-12
2.2 Model Emission Source Development ..... 2-14
2.2.1 Data Gathering 2-14
2.2.2 Model Development 2-17
2.2.3 Model Characteristics 2-27
2.3 Environmental and Energy Impacts of
Controlling Emissions from Process Vents . . 2-27
2.3.1 Primary Air Pollution Impacts .... 2-29
2.3.2 Secondary Air Pollution Impacts . . . 2-29
2.3.3 Other Impacts 2-32
2.4 Cost Impacts of Controlling Emissions from
Process Vents 2-34
2.5 References 2-37
3.0 TRANSFER LOADING OPERATIONS 3-1
3.1 Emission Source Description 3-2
3.2 Model Emission Source Development 3-5
3.2.1 Data Gathering 3-5
3.2.2 Model Development 3-8
VII
-------
Section
Table of Contents
(CONTINUED)
Pacre
3.3 Environmental and Energy Impacts of
Controlling Emissions from Transfer
Loading Operations „ 3-15
3.3.1 Primary Air Pollution Impacts .... 3-16
3.3.2 Secondary Air Pollution Impacts . . . 3-16
3.3.3 Other Impacts „ 3-23
3.4 Cost Impacts of Controlling Emissions from
Transfer Loading Operations ... 3-24
3.5 References 3-26
4.0 STORAGE TANKS 4-1
4.1 Emission Source Description 4-1
4.1.1 Fixed Roof Tanks 4-2
4.1.2 Floating Roof Tanks 4-2
4.1.3 Horizontal Tanks 4-5
4.2 Model Emission Source Development 4-5
4.2.1 Data Gathering 4-6
4.2.2 Model Development 4-6
4.2.3 Model Characteristics 4-13
4.3 Environmental and Energy Impacts of
Controlling Emissions from Storage Tanks . . 4-13
4.3.1 Primary Air Pollution Impacts .... 4-13
4.3.2 Secondary Air Pollution Impacts . . . 4-16
4.3.3 Other Impacts 4-16
4.4 Cost Impacts of Controlling Emissions From
Storage Tanks 4-17
4.5 References 4-21
5.0 WASTEWATER 5_1
5.1 Emission Source Description 5-2
5.1.1 Organic-Containing Wastewater .... 5-2
5.1.2 Air Emissions 5-3
VI11
-------
TABLE OF CONTENTS
(CONTINUED)
Section - Page
5.2 Model Emission Source Development 5-9
5.2..1 Data Gathering 5-9
5.2.2 Model Development 5-10
5.2.3 Model Characteristics 5-11
5.3 Environmental and Energy Impacts of
Controlling Emissions from Wastewater . . . 5-16
5.3.1 Primary Air Pollution Impacts .... 5-18
5.3.2 Secondary Air Pollution Impacts . . . 5-18
5.3.3 Other Impacts 5-25
5.4 Cost Impacts of Controlling Emissions
from Wastewater . , 5-29
5.4.1 Capital Costs 5-29
5.4.2 Annual Costs 5-32
5.4.3 Cost Effectiveness . . . 5-32
5.5 References 5-45
6.0 EQUIPMENT LEAKS 6-1
6.1 Emission Source Description . 6-1
6.1.1 Pumps . . 6-2
6.1.2 Compressors 6-2
6.1.3 Process Valves 6-3
6.1.4 Pressure Relief Devices 6-3
6.1.5 Open-Ended Valves or Lines 6-4
6.1.6 Sampling Connections 6-4
6.1.7 Connectors 6-4
6.1.8 Agitators 6-4
6.1.9 Product Accumulator Vessels 6-5
6.1.10 Instrumentation Systems 6-5
6.2 Model Emission Source Development 6-5
6.2.1 Data Gathering 6-6
6.2.2 Model Development 6-6
6.3 Environmental and Energy Impacts of
Controlling Emissions from Equipment Leaks . 6-11
6.3.1 Primary Environmental Impacts .... 6-11
6.3.2 Secondary Environmental Impacts . . . 6-18
ix
-------
Section
6.4
TABLE OF CONTENTS
(CONCLUDED)
Cost Impacts of Controlling Emissions from
Equipment Leaks
6.4.1 Capital Costs . . .
6.4.2 Annual Costs . . .
6.4.3 Cost Effectiveness
Page
6-22
6-22
6-27
6-28
6.5
References
6-34
-------
LIST OF TABLES
Table Page
2-1 Process Vents Model Stream Characteristics .... 2-15
2-2 Category A Air Oxidation Vent Model Streams . . . 2-20
2-3 Category A Distillation Vent Model Streams .... 2-21
2-4 Category A Reactor Vent Model Streams 2-24
2-5 Category B (Generic) Process Vent Model Streams . 2-26
2-6 Example Model Process Vent Streams . 2-28
2-7 HAP and VOC Emission Reductions for Example Model
Process Vent Streams 2-30
2-8 Secondary Impacts for Example Model Process Vent
Streams 2-31
2-9 Annual Control Cost Estimates for Example Model
Process Vent Streams 2-35
3-1 Transfer Rack Characteristics 3-7
3-2 Model Tank Car Transfer Racks 3-9
3-3 Model Tank Truck Transfer Racks 3-10
3-4 Throughput Allocation. Factors . . 3-14
3-5 Example Model Transfer Racks 3-17
3-6 HAP and VOC Emission Reductions for Example
Model Transfer Racks 3-19
3-7 Secondary Air Pollution Impacts for Example
Model Transfer Racks 3-21
3-8 Annual Control Cost Estimates for Example Model
Transfer Racks 3-25
4-1 Model Tank Farms , 4-8
4-2 Model Tank Types . . . 4-12
4-3 Example Model Tank Farms 4-14
xi
-------
LIST OF TABLES
(CONTINUED)
Table
Page
4-4 HAP and VOC Emission Reductions for Example
Model Tank Farms 4-15
4-5 Annual Control Cost Estimates for Example
Model Tank Farms 4-18
5-1 Emission Sources in Wastewater Collection and
Treatment Systems 5-5
5-2 Model Wastewater Stream Parameters 5-12
5-3 Example Model Wastewater Streams 5-13
5-4 Example Model Wastewater Streams 5-17
5-5 HAP and VOC Emission Reductions for Example
Model Wastewater Streams 5-19
5-6 Combustion Pollutant Emission Factors for Steam
Generation 5-22
5-7 Secondary Air Pollution Impacts of Example Model
Wastewater Streams 5-23
5-8 Annual Fuel Use for Steam Generation for Steam
Stripper Control of Example Model Streams .... 5-28
5-9 Estimation of Basic Equipment Cost for a Steam
Stripping Unit Treating a Facility Wastewater
Flow of 50 £pm 5-30
5-10 Estimation of Basic Equipment Cost for a Steam
Stripping Unit Treating a Facility Wastewater
Flow of 500 £pm 5-31
5-11 Estimation of Total Capital Investment for a Steam
Stripping Unit Treating a Facility Wastewater
Flow of 50 £pm 5-33
5-12 Estimation of Total Capital Investment for a Steam
Stripping Unit Treating a Facility Wastewater
Flow of 500 £pm 5-35
XII
-------
LIST .OF TABLES
(CONCLUDED)
Table
Page
5-13 Estimation of Total Annual Cost for a Steam
Stripping Unit Treating a Facility Wastewater
Flow of 50 £pm 5-37
5-14 Estimation of Total Annual Cost for a Steam
Stripping Unit Treating A Facility Wastewater
Flow of 500 £pm 5-39
5-15 Cost Effectiveness for Example Model Wastewater
Streams 5-42
6-1 Model Unit Equipment Counts 6-9
6-2 Model Unit Parameters 6-12
6-3 Summary of Equipment Leak VOC Emission Factors . . 6-14
6-4 Leak Definitions and Base Performance Levels
for Pumps, Valves, and Connectors 6-16
6-5 Baseline VOC and HAP Emissions for Model Units . . 6-19
6-6 -Model Unit VOC Emissions from MACT Control .... 6-20
6-7 VOC and HAP Emission Reductions for Model Units . 6-21
6-8 Total Installed Base Costs 6-24
6-9 Capital Costs for Equipment for Existing Units . . 6-25
6-10 Initial Monitoring and Leak Repair Cost Estimates 6-26
6-11 Annual Monitoring and Leak Repair cost Estimates
(Monthly Valve Monitoring) .... 6-29
6-12 Annual Control Costs for Model Units
(Monthly Valve Monitoring) 6-30
6-13 Cost Effectiveness for Model Units (Monthly
Valve Monitoring) 6-32
Xlll
-------
LIST OF FIGURES
Figure . page
2-1 General Examples of Reactor-Related Vents v ... 2-3
2-2 Flash Distillation Column 2-6
2-3 Conventional Fractionating Column 2-8
2-4 Potential VOC Emission Points for a Nonvacuum
Distillation Column 2-9
2-5 Potential VOC Emission Points for a Vacuum
Distillation Column Using Steam Jet Ejectors
with Barometric Condenser 2-10
2-6 Potential VOC Emission Points for Vacuum
Distillation Column Using a Vacuum Pump 2-11
2-7 Schematic of a Flowsheet for a Liquid-Phase Air
Oxidation Process 2-13
3-1 Bottom-Loading Tank Truck with Vapor
Collection System 3-4
4-1 Typical Fixed Roof Storage Tank 4-3
5-1 secondary Air impacts from Steam stripper
Control (Controlled Boilers) 5-24
5-2 Comparison of Air Impacts from Steam Stripper
Control (Controlled Boiler) 5-26
5-3 Unit Operating Costs Versus Wastewater Feed Rate
for Steam Stripping Unit 5-41
5-4 Cost Effectiveness Versus Wastewater Feed Rates
for an Example Stream 5-43
xiv
-------
ACRONYM AND ABBREVIATION LIST
AID
atm
BID
Btu
CAA
CMA
CO
C02
CTC
CTG
EPA
ft
gal
Gg
gpm
gpy
HAP
HC1
H/D
HON
hr
kg
kJ
kPa
kW
£
Ib
Ib-mole
LDAR
£pm
additional information document
atmosphere
background information document
British thermal unit(s)
Clean Air Act
Chemical Manufacturer's Association
carbon monoxide
carbon dioxide
Control Technology Center
Control Technology Guidelines
Environmental Protection Agency
foot (feet)
gallon(s)
gigagram(s)
gallon(s) per minute
gallon(s) per year
hazardous air pollutant
hydrochloric acid
height-to-diameter ratio
hazardous organic NESHAP
hour(s)
kilogram(s)
kilojoule(s)
kilopascal
kilowatt(s)
liter(s)
pound(s)
pound mole(s)
leak detection and repair
liter(s) per minute
xv
-------
m
MACT
Mg
min
MMBtu
MMgal
mmHg
MMkj
MMscf
MW
NaOH
NESHAP
NOX
NSPS
OCM
PM
POTW
ppm
ppmv
ppmw
psia
RACT
°R ,
scf
scfm
sec
SIP
S02
SOCMI
VOC
VOHAP
meter(s)
maximum achievable control technology
megagram(s)
minute(s)
million British thermal unit(s)
million gallon(s)
millimeter(s) of mercury
million kilojoule(s)
million standard cubic feet
megawatt(s)
sodium hydroxide
National Emission Standard(s) for Hazardous
Air Pollutants
nitrogen oxide
new source performance standard(s)
organic chemical manufacturing
particulate matter
publicly owned treatment works
part(s) per million
part(s) per million by volume
part(s) per million by weight
pounds per square inch absolute
reasonable available control technology
degrees rankine
standard cubic foot (feet)
standard cubic foot (feet) per minute
second(s)
State Implementation Plan
sulfur dioxide
synthetic organic chemical manufacturing
industry
volatile organic compound
volatile organic HAP
xvi
-------
VOL
wt
yr
$
volatile organic liquid
weight
yearns)
dollar(s)
xvn
-------
-------
1.0 INTRODUCTION
In developing a national emission standard for
hazardous air pollutants (NESHAP), the U. S. Environmental
Protection Agency (EPA) must assess the baseline emissions
from a source category and the impacts of various control
options for regulating that source category. Impacts of the
control options include reduction in emissions of the
regulated pollutants, cost of implementing the control
requirements, changes in energy consumption, and increases
or decreases in emissions of other pollutants indirectly
affected by the options.
To quantify the national impacts of the options
available for regulating emission sources in the synthetic
organic chemical manufacturing industry (SOCMI), impacts of
the options could be examined for each individual SOCMI
facility in the country. However, for the hazardous organic
NESHAP (HON) the detailed information needed for such an
assessment was not available for each facility and gathering
such data could not be accomplished if the promulgation date
of within 2 years of enactment of the Clean Air Act was to
be met. The similarity in operations at SOCMI facilities
did, however, allow the use of model emission sources to
represent actual emission sources at the various facilities.
This approach resulted in estimates of impacts for typical
SOCMI facilities which can, in turn, be extrapolated to
estimate impacts for the SOCMI on a national level.
This volume presents the methodology used to develop
model emission sources for the HON. The emission source
types described here are common to SOCMI facilities
throughout the country:
• Equipment leaks;
• Process vents;
• Storage tanks;
1-1
-------
• Transfer loading operations; and
• Wastewater collection and treatment operations.
Within this volume, each chapter describes the data-
gathering efforts and model development procedures used for
one of these five emission source types. Also presented are
summaries of baseline emissions, emission reductions,
control costs, and secondary emissions for several example
models.
The appendices to this volume contain example
calculations of baseline emissions, emission reductions,
cost effectiveness, and secondary air pollution impacts.
Detailed cost calculations for each of the control
technologies used in the HON analysis were presented in the
appendices to Volume IB of this background information
document (BID).
1-2
-------
2.0 PROCESS VENTS
Reactor processes, distillation operations, and air
oxidation processes in the SOCMI have the potential to emit
volatile organic compounds (VOC's), many of which are organic
hazardous air pollutants (HAP's), from process vents.
Emissions from process vents may be released either directly
to the atmosphere, downstream of a recovery system, or to a
control device that is vented to the atmosphere. Emissions
from process vents can be reduced through the use of an add-on
recovery or combustion device. Although a variety of control
devices are suitable for the different process vent streams in
the SOCMI, the most universally applicable control technique
is combustion.
This section discusses potential emission sources, models
used to represent process vents", and impacts of the control
technologies for process vents in the SOCMI. Section 2.1
presents a brief description of the sources of process vent
emissions. Section 2.2 discusses the model process vents
developed to represent vent stream characteristics, complete
with emissions of VOC's and organic HAP's. The environmental
and energy impacts of controlling emissions from process vents
are presented in Section 2.3. The cost impacts of the control
technologies are given in Section 2.4.
2.1 EMISSION SOURCE DESCRIPTION
To manufacture organic chemicals, a process unit may use
technologies from two broad categories of processes:
conversion and separation. Conversion processes involve
chemical reactions that alter the molecular structure of
chemical compounds. These processes are included in the
reactor and air oxidation process segments of the SOCMI.
2-1
-------
Separation processes often follow conversion processes and
divide chemical mixtures into distinct fractions such as
products, by-products, reactants, and such. Distillation,
stripping, absorption, filtration, crystallization, and
extraction are all separation processes. Distillation is the
separation process addressed in this analysis, because it is
the most widely used separation process and has the potential
to release larger amounts of VOC's and HAP's than other
separation processes.
The potential organic HAP and VOC emission sources from
reactor processes are discussed in Section 2.1.1 of this
section, distillation operations are discussed in
Section 2.1.2, and air oxidation processes in Section 2.1.3.
2.1.1 Reactor Processes
In this study, the term "reactor process" refers to means
by which one or more substances, or reactants, are altered by
any chemical reaction other than air oxidation, so that one or
more new organic chemicals are formed. The air oxidation
process is considered to be a separate type of process vent
because air or air enriched with oxygen acts as the oxidizing
agent, resulting in larger-volume reactor vent streams and
thus, potentially higher VOC emissions.
Emissions of VOC's can be released from reactor process
vent streams and from product recovery systems associated with
reactors. Product recovery equipment includes condensers,
absorbers, and adsorbers, which are used to recover products
or by-products for use, reuse, or sale. Product recovery
equipment does not include product purification devices
involving distillation operations.
Reactor processes may involve either liquid-phase
reactions or gas-phase reactions. Four potential atmospheric
emission vent types are shown in Figure 2-1 and include the
following:
(A) Direct reactor process vents from liquid-phase
reactors.
(B) Vents from recovery devices applied to vent streams
from liquid-phase reactors. (Raw materials,
2-2
-------
Liquid-Phase Reactor
Gas
Vent Type A
Liquid
Gas
Product/By-Product
Recovery Device
Vent Type B
Recovered
Product
Liquid
Gas-Phase Reactor
Process Vents Controlled by Combustion
Process Vent Streams
from A, B, or C
Gas
Vent Type C
Liquid
Gas
Vent Type D
Combustion
C
Figure 2-1. General examples of reactor-related vents.
2-3
-------
products, or by-products may be recovered from vent
streams for economic or environmental reasons.)
(C) Process vents from gas-phase reactors after either
the primary or secondary product recovery device.
(Gas-phase reactors always have primary product
recovery devices.)
(D) Exhaust gases from combustion devices applied to any
of the above streams.
Some chemical processes may not be vented to the atmosphere at
4
all, while other processes may have one or more vent streams.
The characteristics of reactor vent streams in the SOCMI
vary widely among the numerous chemicals and chemical
reactions. Process vent streams show a great variety in heat
content, volumetric flow rates, chemical compositions, and VOC
and HAP concentrations. In addition, the possible combination
of product recovery devices and reactor processes introduces
an additional source of variability among emission
characteristics from the similar reaction types.
2.1.2 Distillation Operations
Distillation separates one or more feed streams into two
or more outlet streams, which have component concentrations
different from those in the feed streams. Separation is
achieved by redistributing the components between the liquid
and vapor phase as they approach equilibrium within the
distillation column. The more volatile components concentrate
in the vapor phase and the less volatile components
concentrate in the liquid phase.
Distillation systems can be distinguished according to
the operating mode, the operating pressure, the number of
distillation stages, the introduction of inert gases or steam,
and the use of additional compounds to aid separation. A
distillation unit may operate in continuous or batch mode, and
at operating pressures (1) below atmospheric (vacuum),
(2) atmospheric, or (3) above atmospheric (pressure).
Distillation can occur as a single stage or multistage
process. To improve separation, inert gas or steam is often
introduced. Steam is often sparged into the bottom of the
2-4
-------
distillation column as a substitute for a reboiler. Finally,
compounds are often introduced to aid in distilling mixtures
containing constituents that are hard to separate (i.e.,
extractive distillation).
Single-stage batch distillation is not common in large-
scale chemical production but is widely used in laboratories
and pilot plants. Batch distillation is also used in organic
intermediate manufacturing. Separation is achieved by
charging a still with material, applying heat, and
continuously removing the evolved vapors. In some instances,
steam is added or pressure is reduced to enhance separation.
Single-stage continuous distillation is referred to as
flash distillation. Figure 2-2 illustrates a simplified flash
distillation column. It can be generally defined as a direct
separation of a component mixture based on a sudden change in
pressure. Because it is a rapid process, steam or other
components are not added to improve separation. Flash
distillation is frequently the first separation step for a
stream coming from a reactor. The heated products from a
reaction vessel are transferred to an expansion chamber. The
pressure drop across the valve, the upstream temperature, and
the expansion chamber pressure govern the degree of separation
achieved. The light ends quickly vaporize and expand away
from the heavier bottom fractions, which remain in the liquid
phase. The vapors rise to the top of the unit and are
removed. The bottom fractions move on to the next process
step.
Fractionating distillation is a multistage distillation
operation. It is the most commonly used type of distillation
process in large organic chemical plants, and it can be a
batch or a continuous operation. Fractionating distillation
is accomplished by using trays, packing, or other internals in
a vertical column to provide intimate contact between
ascending vapors and descending liquid streams. Concentration
gradients in the vapor phase and liquid phase are achieved
across the length of the column. A simplified block flow
2-5
-------
Overheads (Gas) or
Light Ends
Pressure Control
Valve
Feed
Flash Distillation
Column
Bottoms (Liquid) or
Heavy Ends
Figure 2-2. Flash distillation column.
2-6
-------
diagram of a fractionation column is shown in Figure 2-3. The
light-end vapors evolving from the column are condensed and
collected in an accumulator vessel. In a fractionating
column, part of the distillate (i.e., condensed light ends) is
returned to the top of the column so it can fall
countercurrent to the rising vapors. For difficult
separations, a chemical agent is added to change the
properties of the mixture and thus facilitate the separation.
This is referred to as extractive distillation. Another
distillation technique is desorption. A desorption column is
very similar to a fractionating distillation column except
that it does not use a reflux condenser.
Vapors separated from the liquid phase in a distillation
column rise out of the column to a condenser. The gases and
vapors entering the condenser can contain VOC's, water vapor,
and noncondensibles, such as oxygen, nitrogen, and carbon
dioxide (062). The vapors and gases originate from the
vaporization of liquid feeds, dissolution of gases in liquid
feeds, addition of inert carrier gases to assist in
distillation (only for inert carrier distillation), and
leakage of air into the column, especially in vacuum
distillation. The gases and vapors entering the condenser are
cooled and the condensible gases are collected as a liquid.
If present in the condenser, noncondensible gases, such as
oxygen, nitrogen, CC>2, and other organics with very low
boiling points, do not usually cool sufficiently to condense
and, therefore, are emitted in the vent stream from the
condenser. Portions of this vent stream may be recovered with
additional control devices such as scrubbers, adsorbers, and
secondary condensers. Vacuum-generating devices, such as
vacuum pumps and steam .ejectors, might also affect the removal
of noncondensibles from a vent stream.
Emissions of VOC's may be released when noncondensible
gases containing some hydrocarbons are vented. The most
frequently encountered emission points from fractionating
distillation operations are illustrated for several types of
distillation units in Figures 2-4 through 2-6. These emission
2-7
-------
Residue
(Bottom Products)
Feed »•
V = Vapor
LI l/ii iM
- f \
/^
T
V2
t
V3
t
V4
t
V5
t
V6
t
V7
t
V8
t
V9
I Heboiier J
J, .— f—\X^ ^X
^
1
L1
I
L2
I
L3
I
L4
I
L5
I
L6
I
L7
i
LO
— »f Reflux ^\ — "-Coolant
y^ Condenser )
( Accun
\ Tr
V Tc
nulator ^
nk J
i r
Heating Medium
(Bottom Products)
Distillate
(Overhead Products)
Figure 2-3. Conventional fractionating column.
2-8
-------
Vent to Atmosphere
Pressure Relief
Valve (6)
Overhead Product
Distillation
Column
Figure 2-4.
Potential VOC emission points for a nonvacuum
distillation column.
2-9
-------
Steam
Steam Jet
Ejector (4)
Cooling
Water (CW)
Steam
Steam Jet
Ejector (4)
Overhead Product
Distillation
Column
Vent
Wastewater
Vent
,Wastewater
Figure 2-5. Potential VOC emission points for a vacuum
distillation column using steam jet
ejectors with barometric condenser.
2-10
-------
Vapor Phase
C
Liquid Reflux
Distillation
Column
CW
Condenser (1)
Vent
Vacuum Pump (5)
Accumulator (2)
Overhead Product
Figure 2-6. Potential VOC emission points for a
vacuum distillation column using a vacuum pump.
2-11
-------
points include (1) condensers, (2) accumulator vessels,
(3) hot wells, (4) steam jet ejectors, and (5) vacuum pumps.
The total volume of gases emitted from a distillation
column depends on the volume of air leaks into the vacuum
column, volume of inert carrier gas used, volumes and types of
gases dissolved in the feed, efficiency and operating
conditions of the condenser or other process recovery
equipment, and physical properties of the organic
constituents. Knowledge of the quantity of air leaks and
dissolved gases in the column used in conjunction with
information on physical properties in the organic vapor and
condenser operating parameters allows estimation of the VOC
emissions that may result from a given distillation
, . 7
operation.
2.1.3 Air Oxidation Processes
In an air oxidation process, oxygen in air reacts with an
organic compound to introduce one or more oxygen atoms into
the compound, to remove hydrogen or carbon atoms from the
compound, or a combination of both. An example air oxidation
process is illustrated in Figure 2-7. The air oxidation
process is a subset of reactor processes in which air or air
Q
enriched with oxygen is the oxidizing agent.
Air oxidation processes vent large quantities of inert
materials containing usually low concentrations of VOC's to
the atmosphere. These inerts, predominantly nitrogen, are
present because air contains 20.9 percent oxygen.and
78.1 percent nitrogen by volume on a dry basis. The nitrogen
in the air passes through the process unreacted. The quantity
of nitrogen and unreacted oxygen emitted to the atmosphere is
a function of the amount of excess air used in the air
oxidation process.9 The characteristics of air oxidation vent
streams vary among the different chemical production processes
in heat content, volumetric flow rate, composition, and
concentration of VOC's and HAP's.
2-12
-------
c
'•go
-
CD
og.
o-
w
(0
0)
o
o
c
o
•H
4J
•H
X
o
•H
(0
0)
M
(0
A
A
•O
•H
•H
rH
(0
O
-p
0)
0)
X)
U)
g
O
•H
-P
(0
,
o
w
i
CM
0)
2-13
-------
2.2 MODEL EMISSION SOURCE DEVELOPMENT
To assess the national impacts of control options for
process vent air emissions from SOCMI process units, the
impacts of the options on all SOCMI process units should be
examined. However, because the detailed information for such
an assessment was not available for each process unit in the
country, baseline emissions and impacts for each process unit
were estimated using model process vent streams to approximate
the process vent streams from different production processes
found at SOCMI process units.
Model process vent streams were developed to represent
the broad range of production processes as well as the range
of emissions common in the industry. Existing data from
previous new source performance standards (NSPS) for the SOCMI
were reviewed to develop the characteristics for model process
vent streams.
The procedure used to develop these vent stream
characteristics is described in the following sections.
Section 2.2.1 discusses how the data were gathered, and the
development of the models is discussed in Section 2.2.2.
Example vent streams are presented in Section 2.2.3.
2.2.1 Data Gathering
During the mid 1980s, NSPS were developed for process
vent emissions from SOCMI distillation operations, reactor
processes, and air oxidation processes. For each standard, a
BID1'3'6 was developed that characterized the vent streams
associated with each process vent type, in terms of flow rate,
heat content, and VOC concentration of each stream. These
vent stream characteristics were developed from Section 114
questionnaires gathered under the authority of Section 114 of
the Clean Air Act that focused on emissions of VOC's.
For the HON analysis, the data presented in the SOCMI
NSPS BID's were used to develop three data bases — one for
each respective vent type. ' ' The information stored in these
data bases is described in Table 2-1. For most streams, the
distillation and reactor BID's contain data for process
2-14
-------
TABLE 2-1. PROCESS VENTS MODEL STREAM CHARACTERISTICS
1) Model stream name
2) Reaction type (if applicable)
3) Process capacity corresponding to the model stream
(MMkg)
4) Number of distillation columns (if applicable)
5) Column operating conditions (if applicable)
6) Model stream flow rate (scfm)
7) Model stream heat content (Btu/scf)
8) Model stream temperature (°F)
9) Model stream oxygen content (volume percent)
10) Model stream VOC composition (weight percent)
11) Model stream HAP composition (weight percent)
12) Annual VOC emission rate (kg/yr)
13) Annual HAP emission rate (kg/yr)
2-15
-------
capacity, volumetric flow rate, heat content, stream
temperature, oxygen content, and VO.C mass flow rate. The
total stream mass flow rate was calculated using the ideal gas
law as follows:
10
VOC Weight Percent =
VOC Mass Flow Rate (Ib/yr)
Total Stream Mass Flow Rate (Ib/yr)
The air oxidation NSPS BID only provided information on
stream flow rate, VOC mass flow rate, and heat content. A
separate summary11 of the original Section 114 questionnaire
, .. j_ 12,13,14,15,16,17,18,19,20,21,22
data and the actual responses
were used to identify process capacity, stream temperature,
and oxygen content. The ideal gas law assumptions used to
calculate VOC weight percentages for reactor and distillation
streams were then used to calculate VOC weight percentages for
air oxidation streams.
Information on HAP content was not provided in the NSPS
BID'S, so the Section 114 questionnaires used to develop the
BID'S were examined. For reactor and distillation vents,
speciated vent stream composition was provided in a summary of
the Section 114 responses.23 The overall HAP emission rate
for each stream was estimated by summing the emission rates of
each HAP compound reported for the vent stream.
Stream composition information in the air oxidation
Section 114 questionnaire summary was presented in terms of
feed stock, product, and other VOC's. For example, a stream
might be 2 percent unreacted feed stock, 5 percent product,
and 2 percent other VOC's with the balance being oxygen,
nitrogen, water, carbon monoxide (CO), and 002• It was
assumed that "other VOC's" were HAP's. The mass flow rate of
HAP's was estimated by adding the flow rate of "other VOC's"
to the flow rates for the feed stock (if it was a HAP) and
product (if it was a HAP).
2-16
-------
For reactor, distillation, and air oxidation vent
streams, stream HAP weight percentage was calculated as
follows:
HAP Mass Flow Rate _ HAP Weight Percent
VOC Weight Percent
VOC Mass Flow Rate
or
HAP Weight Percent =
/HAP Mass Flow Rate\
\VOC Mass Flow Rate/
* VOC Weight Percent
For some streams, HAP weight percentage could not be derived
from the available data. For these streams, the following
equation was used:
HAP VOC
Weight = Weight * /Average HAP Weight Percent\
Percent Percent \Average VOC Weight Percent/
where the average HAP and VOC weight percentages were
calculated as the mean of the values for all streams for which
HAP data were available. The HAP mass flow rate was
calculated as follows:
10
HAP -VOP
Mass flow = Mass Flow * (HAP Weight Per cent \
Rate Rate \VOC Weight Percent/
Some streams in the SOCMI NSPS BID's lacked data for
stream temperature or stream oxygen content. When such data
were not available from the Section 114 responses, average
values were used. These average values were calculated as the
mean of the values for all streams for which the data were
available.
2.2.2 Model Development
The process vent data bases were used to generate model
streams to characterize uncontrolled emission streams or
emission streams located upstream of combustion devices. The
models represent vent streams from reactor processes,
distillation operations, or air oxidation processes. Model
2-17
-------
stream characteristics were developed for each vent type for
specific production processes and for generic process types.
The production process-specific streams, or Type A models,
represent an actual production process, as described in the
SOCMI NSPS BID's. The generic model streams, or Type B
models, represent a general classification of the production
processes for which specific information was not presented in
the SOCMI NSPS BID's, such as a model for a process using a
halogenation reaction.
2.2.2.1 Type A Uncontrolled Models. The completed data
bases were used to generate the Type A model streams for the
specific production processes identified for each vent type.
If only a single stream was identified for a particular
production process, its characteristics were designated as the
model. If multiple streams were identified for a particular
process, then a single set of vent stream characteristics were
developed based on median characteristics, weighting by
capacity, or weighting by flow rate.
The vent stream characteristics for each Type A
production process were derived from data on all existing
production process streams presented in the SOCMI NSPS BID's.
The percent oxygen represents the median oxygen value. The
stream temperature represents the median temperature value.
The model capacity represents the median of all the
capacities associated with vent streams for that process. For
example, data were available for five reactor vent streams
from the production of cumene by alkylation. After sorting
the streams from lowest to highest capacity, the capacity of
the third stream would be the median capacity and, therefore,
the capacity for the model for that production process.
10
2-18
-------
Model flow rate was calculated as presented in the
following equation:
Model Flow Rate = Median Capacity ^
N
S Capacity
where median capacity is for stream i=l through N. The number
of columns, VOC emissions, and HAP emissions were calculated
with the same approach, by replacing the sum of the flow rates
with the sum of the respective characteristics.
Model heat content and VOC weight percent were calculated
as weighted averages based on flow rate. The model heat
content was calculated as shown in the following equation:
N
Model Heat Content = S
' (Heat Content) i * (Flow Rate) iN
N
S
Flow Rate
A similar calculation was used to calculate the weight percent
VOC and the weight percent HAP. The weight percent VOC was
calculated as shown:
N
Weight Percent VOC = 2
' (Weight Percent VOC) i * (Flow Rate) i'
N
Flow Rate
. Model vent stream characteristics for air oxidation
processes are presented in Table 2-2 and distillation
operation model vent stream characteristics are presented in
Table 2-3. Because occasionally a chemical is produced using
both an atmospheric column and a vacuum column, data for the
characteristics of both types of streams were combined to
represent a single distillation operation. In addition, the
number of operating columns for each model was developed to
estimate impacts per operating column. The model vent stream
2-19
-------
CO
w
EH
CO
H
Q
O
s
tH
w
>
•5
0
EH
Q
H
X
O
p/
H
<
o
w
EH
fCQ
U
•
1
OJ
TABLE
C
O
*•* •£?
Q_ •-- *^
= |s
u
oJ«£
^ CO (D "*"*
*e ^ z
LLJ ^
g
o •— **
> |s
U O
T ° c
O > ,° 4) ^
« "ra « 2 oi
jr j-> '= "^ z
5 £ u3 ^
cu C r- O
O
O (j v^
§ o
1— ^
^^
« il
= gs
o ra
o
o
CM
in
ro
•o
o
O>
e>
3
as
(U
?.
lo
(0 C
4-> ID
8} .C
U *-*
-
0
CD
CM
0>
^>
CM
*~
o-
in
CM
o
o
g
in
o
c
0)
II
[_
lonitrile
nia and p
>* o
1 1
£_
^
C
o
">.
c.
u
<
o
o
o
e
'•O
03
o
CD
CO
CM
0
vf
e
CM
CX
in
o
0
CM
CM
•O
s
(D
>g
TJ *-*
— (U
X TJ
°X
g°
— • O
^- 0)
LU TJ
TJ
X
O
01
I
'x
4->
UJ
,_
CD
0
O
«*
o
5
o
•o
o
s
~a-
o
in
CM
CO
•o
1
TJ
e
O
01
•R
-g
TJ
g
L.
O
u_
0.
CD
O
in
""
CM
O
0
in
ro
ro
~
o
o
'^
^
0
CD
O
O
ro
CM
CO
10 0)
'> g
N
51
L.
^L*o
•S c
(D O
fil "D
(0 X
•o
f
c
(D
U
0>
<0
z
y_
CD
O
CM
O
ro
ro
o
CD
CM
3
^
CM
"
ro
«r—
§
CM
o
1
5
o-
CD
01
TJ
ID
(D
-C
j:
(U
TJ
t.
TJ
>«
ro
o
ra
•M
j:
a.
CM
CD
0
R
CM
O
CD
£
CM
ro
in
o
s
o
in
r-^
0
CD
s
o
o
CM
TJ
as
phthalic
t-
0)
TJ
'o
(D
U
-C
1.
0)
I_
-------
CO
w
Q
O
S
w
o
H
EH
EH
CO
H
Q
O
O
W
I
(N
CQ
<
EH
CO
u
4->
L_
CD
u
t_
ro
x:
o
E
CD
£_
*J
C/)
C
0
Q. .,_ 5x
>"
.^ to c
^ a: z:
c
o
O « 4_,
o
§ o e
_ 'w S ^*
10 .^ S ra
o .5 «—•
1 —
-M x-*
|s^
x o >
O cj ^
Q._
H- *"'
4-* N-
= §5
CJ CO
3
|
Ou
in
o
r\J
0
f\j
o
ro
CM
CD
5
*~
CD
in
CM
J3
-J-
3
CO
in
ro
O
in
CM
=
^
T3
U
CO
U
4-t
CU
O
0
O
o
o
0
o
CD
0
O
CD
0
O
O
CD
0
CV1
CO
0
o
CD
CD
in
f-
co
•o
>
o
snitrile
a.
is
0
CD
O
O
CD
O
CD
O
O
CD
0
0
0
O
0
^
0
0
CD
0
CD
O
CO
>o
^
ro
I alcohol
>^
5
CD
0
o
o
CD
0
O
o
o
CD
O
CD
O
CD
O
O
CD
0
0
0
o
CD
CD
O
in
CM
>
g
CM
I chloride
>H
S
CD
O
in
«-
CD
0
O
in
«-^
s
o
(M
CO
0
fM
ro
s
o
CO
^o
>
CM
C
,
'E
0
o
*-
'
CD
0
•-
^
O
o
CD
j:
O
'in-
"~
CM
0
•z.
ro
0)
I
CO
3
CO
o
0
o
0
o
0
o
o
o
0
o
o
o
o
o
o
0
o
CD
o
o
o
0
3
...
1
^
E
4-1
u
CO
o
c_
Q.
CO
o
o
o
CD
O
0
o
0
o
0
CD
0
0
0
CD
0
0
o
CD
§
O
CM
CM
=
CM
CJ
an disulfid
.5
u
CO
u
0
o
CD
O
O
O
O
0
0
o
0
0
0
o
0
0
0
CD'
0
CD
CD
»
O
=
IO
~o
ro
o
CO
o
o
^
=
^_
'obenzene
o
5
0
CD
CD
0
O
0
CD
O
O
CD
O
O
O
0
O
O
0
CD
O
CD
O
CO
*o
=
•0
0)
c
CO
x:
cu
*odif luorom
o
-C
CJ
o
CD
O
o
0
o
0
o
o
CD
O
o
CD
O
O
O
O
O
o
CD
0
g
CNJ
=
^
CJ
CJ
CJ
CD
0
o
o
0
o
o
0
CD
0
CD
O
O
CD
O
O
CD
CD
O
0
o
ro
£
=
V
)hexane
— *
u
o
OJ
CM
h~
N.
r~
^
in
o
ro
o
•o
o
o
0
CM
C>
CO
o
in
•o
CO
5
=•
CO
c
CO
X
01
.c
o
11
)hexanone/c'
.
CJ
CJ
§
o
0
CD
O
0
o
o
o
0
0
0
o
0
^
o
0
CD
0
CD
O
CO
-o
=
CM
CU
c
CO
.c
CM
CJ
4-1
CO
CO
|
0)
-C
flj
1
o
0
o
o
0
o
CD
O
o
0
o
o
CD
in
o
o
CD
o
CD
o
o
*3
CO
^
^
2
ro
to I amines
^
, acrylate
x:
4-»
UJ
sO
in
in
ro
T—
O
*t
in
in
ro
*~
O
>3-
«-
in
0
in
*-
sT
CD
fM
^
in
.acetate
01
e
CM
ro
^
o
CM
ro
,_
0
o
o
«-
CM
CM
ro
o
CM
o
in
"~
>
2:
CM
.benzene
JZ
UJ
o
0
o
0
CD
O
CD
0
0
0
o
o
CD
O
O
0
o
CD
CD
CD
CO
in
CM
S
*
>
^
ro
"o
D)
OJ
jr
4-*
LU
CO
"-
o
ro
CD
CO
-J-
*~
0
10
0
o
o
o
o
CD
o
CD
*
0-
^>
o
SS
>
^.
i
1
E
o
u_
0
CD
CD
O
O
O
0
o
CD
0
0
CD
O
O
<-
O
CD
CD
O
0
O
eg
'
=
CU
g
"^
1
^
cc
o
in
ro
in
o
o
0
•o
in
CM
•o
^>
§
T—
O
CM
L5
s:
1
CD
i—
0
s:
2-21
-------
w
04
EH
W
P
w
o
H -—
EH Q
t—3 Q
H J
EH U
H O
Q O
< ^
CD
w
I
P3
IS
to
U
4-*
CO
L.
4-*
t_
CO
CJ
01
K
g
«< CO ,,
DC O 5
E ^"'
O
C
o
'w *r
,«.
^ flj D)
Q. ^
3;
C
o
O CO ^j
g- v^
O
CJ
o g ~
__. CO -M ^
«.|S »
1—
c cQ
|«-
&~
" ^
4-« H-
., C 0
ra S «
O 4-J
CJ CO
3 <1) E
— • (0 C
u_ o: co
£££
o o ^
o
0
o
t>
V*
CO
OJ
CD
ro
0
r*-
o
•o
ro
o
5?
z
OJ
a
g
cu
.g
O «-
in
O^ CO
«r- *~
| >
to «-
•M
CO
0
vj- o
CO CD
«—
OJ O
CO CD
ro o
0
*"dr o
CO O
*~
OJ CD
CO 0
ro o
0 0
O CD
0 O
O
CD
-4- O
*~"
O 0
ro o
«- o
CO O
NO in
z i
OJ >O
~o
O <1>
ca c
CO
o _c
— ' 0)
ca o
I -2
(U U
L_ ••—
0) L.
\— 1—
O
0
0
CD
0
O
o
o
o
0
o
o
CD
0
0
0
CD
O
0
CD
0
O
s
•Si
ro
a>
0)
">*
-M
0)
o
o
U
£_
h-
O
O
o
o
o
o
CD
0
o
o
o
o
o
0
o
o
CD
O
o
CD
O
CO
O
^
r-
CD
J=
*-*
Q)
O
t_
o
H-
o
o
o
1—
t-
m
0
•*3"
T-
ro
i —
in
0
^*
*-
P2
o
o
o
f T—
in
"*
o
o
s
1
•*
0)
(0
•M
0)
U
CD
C
•f—
"*
o
0
CD
O
O
O
0
o
o
0
o
o
0
o
o
o
c:
CD
O
C
o
OJ
CO
z
ro
T
L.
c
(.
c
•*~
"^
ii ii
> >
2-22
-------
represents one combined vent stream for all distillation
columns used in the process.
The specific model vent stream characteristics for
reactor processes are shown in Table 2-4. For reactor
processes, the production process for the model is specific to
the actual reaction type for each product. For example, two
sets of model vent stream characteristics were developed for
linear alkylbenzene. The first model stream represents the
neutralization reaction, and the second model stream
represents the aikylation reaction.
2.2.2.2 Type B Uncontrolled Models. Type B models, or
generic models, were developed to represent production
processes with no specified stream information in the SOCMI
NSPS BID'S. The Type B model streams were developed using the
same approach presented in Section 2.2.2.1 for development of
Type A models. However, the stream characteristics were not
based on specific production processes, but on a range of
processes.
Table 2-5 lists the 22 Type B model streams. A single
Type B model stream was developed for air oxidation processes.
All air oxidation vent streams were considered in developing
this model stream.
For vent streams from distillation operations, two Type B
model streams were developed—one to represent atmospheric
operations and one to represent vacuum operations. All
streams in the updated SOCMI data base were separated by
operating conditions (i.e., atmospheric and vacuum) before
applying the approach from Section 2.2.2.1.
For reactor processes, Type B model streams were
developed based on reaction type. The updated SOCMI data base
2-23
-------
a
"
•~
.<"
[_
0
*-•
s
u
CO
JC
CJ
E
10
G)
l_
4-»
CO
I
< *frt
C/l jj
cf
4I2I
w ~
§
cjS*
§ o w
5"^
5
<->
> 0 17
0 S S
£ Ul ~
^ C 3-5
X* C °
8-^
So
1- ^
*" 4J ^x
= g 5
u m
X 01 E
O 4-» **-
— • CO O
U- OC W
o o ^"
111
I- CJ ^
Q-
!
g
4-t
U
(D
(U
o:
u
•g
o
L.
O.
•3-
O
^
ro
CO
>o
in
o
r—
o
CO
o
CO
"*"
o
°
0
o
N-
co
g
(0
Hydrodimei
0)
L.
4-1
Q-
i
0
o
o
0
o
o
CD
O
O
o
o
0
0
o
o
«
0
o
o
o
o
o
;*
§
Halogenati
Ol
TJ
C.
Q
1
>•
<
O
0
o
o
o
o
CD
O
O
o
o
o
o
o
o
CD
CO
•o
o
o
o
o
0
o
s
i
Hydrogenal
S
^
1
j:
0
N.
in
CM
in
o
o
ro
ro
CO
o
CO
^
s
CM
0
JS
\f\
R
0)
c
'E
c.
o
H-
s
<
»
4-*
u
ID
o
L.
0.
CO
o
0
o
CD
o
o
o
o
CD
o
o
o
o
0
o
o
CD
«
g
o
CD
0
o
CM
ro
g
JLJ
Sulfurizal
2
•4-
3
W)
C
.8
ro
u
CO
CM
*-
^.
CM
•O
«-
O
in
CO
o
CO
o
o
o
o
in
in
R
c
o
Halogenat
01
c
N
1
O
t—
O
jc
o
CD
0
0
§
o
0
o
o
o
o
o
o
0
0
o
o
o
o
o
o
CD
s
c
Alkylatiol
4)
i
CJ
CD
o
o
o
CD
o
o
o
o
o
o
o
0
CD
o
CD
«
o
o
o
CD
o
o
R
o
c
o
Hydrogena
s
s
a
o
""
0
CD
CD
O
o
CD
O
CD
O
O
o
o
0
o
CD
CD
CO
CD
O
O
o
o
o
CM
ro
c
o
ffl
Esterific
o
as
CO
e-
t
at
4-*
JC
i
o
o
o
o
o
CD
O
o
o
CO
o
CO
CD
O
o
CM
CM
CO
O
O
Nitration
Ol
1
o
4-t
o
'c
o
g
0
o
0
o
o
o
o
o
o
o
o
0
0
0
o
o
o
o
CD
o
in
r^
(/>
(/I
CO
1
I
UJ
o
e
^
fc
o
CD
in
V)
o
0
0
o
CO
o
CM
O>
in
CM
^
in
g
4-1
S
Ol
en
Hydrohalo
CM
CO
CD
CO
ro
rC
;*
ro
c
o
Halogenat
at
T3
dichlor
0)
JC
UJ
o
o
CD
O
O
CD
O
0
o
o
o
o
o
0
o
U
O CO
— Ol
4-*
Ol Ol
J=~£
UJ
o
CD
CD
g
0
o
o
CD
O
o
o
0
CD
0
s
>o
0
o
o
o
CD
CD
CM
ro
g
Halogenat
^
m
ro
CM
CM
CM"
o -
Ol «-
L. V
CD
0
0
o
0
CD
O
0
CD
O
O
o
o
o
o
o
o
o
CD
0
0
CM
ro
c
Alkylatio
Ol
c
Ol
N
1
|
to
to
s
-•
ro
CM
CO
"*
f-
ro
C\J
co
«3-
g
^.
CD
O
in
o
c
0-
c
c
c
r<
C
a
O)
Hydrohalo
2
L.
C
JZ
u
^
a:
2-24
-------
w
§
£/j .
rV
S
w
-1
w
Q
o
s
EH
W
o o"
EH W
U Q
W HH1
OH U
Is
«8
^4 x— *
P^
O
a
w
EH
U
•
1
W
ffl
EH
S
rf W) ,
Is
U
c _
O _^J
a. '!T JU X
^ » <0 "ro
il ^ ~
§
*•* 5?
O eft .,
•- t
M
r " c ^
01 5 o ^j
tj *W +J ^
2 3 | S »
CJ t~"
E ., ^
ca c 7: JJ
0) 0,0)
" g 0 >
i'^
Scs"
1— v^
^
jj C t]
-P- C 3
O *^
CJ CD
3
o
£_
"§_
Ol
o
Ol
c
o
01
-^
4-1
01
"x
4^
z
o
0
CD
o
0
o
o
CD
O
CD
O
CD
O
CD
O
O
CO
o
1C
CO
0
CO
^
3
o
r*>
«~
CO
in
g
(U
•M
~
Ol
. c
01
1
o
U
X
0
0
o
CD
O
o
o
o
o
o
o
0
o
o
o
o
CO
<0
o
o
CD
CD
O
CD
CM
CM
C
O
4J
CO
"x
"*
1
Ol
•s.
1
o
X
CD
O
CD
O
O
0
o
0
CD
CD
O
O
O
O
o
CD
O
o
o
o
o
CD
O
O
N.
CM
CM
01
O)
CO
__
O
o
c
o
4-*
01
U
CO
o
c
Ol
.c
a.
£
CM
O
CM
CM
If
in
'
CO
CM
in
in
co
o
o
o
"*•
0
§
o
0
in
o
fO
c
o
O)
0
CQ
o
t
.c
(U
a
41
?*-
£_
O
JC
U
c
2
|
>
2-25
-------
u
c^
CQ
U
in
w
PQ
CO
,0
CO
"u
s.
%
L.
CO
6
i
a
j_i
U)
1
•x.
c
o
a. •— **
^ to . .
&
i^
•*- >•
(/> C71
••- ^E.
01 0)
a. 4J
|
s-st!
u
§ ° ^
_, '« z
>— "* n
c^
11!
Us
0-
5J o
»— vx
^^
C *O
= S^
U CQ
gfl) E
4-* M-
— • (0 O
U. O£ C/
1.^
Ill
t_ o ^
a.
g
ft
a
u
l°l
5 c
z CJ
"1
1—
CO
CO
a
o
0
c.
a.
II
CO
o
ON
in
NO
ON
O
I
in
o
si
o
CM
O
g
t\J
id
-
<
Air Oxidation
«c
0 CM
to o
NO
N± NO
CM
•r- ON
in «—
N.
CM CM
in «—
O st
0 ON
o to
CM
r- ON
«- CO
g %
-4- sf
CM NO
tO CM
NO in
CM 00
- >
CM CM
Distillation M
Distillation V
0 O
CO 0
•* o
*—
" ro
CM
N* O
^
in co
ON
CM
O CO
0 T-
o o
0 0
in in
CM O
*- CM
o
CM O
0
CD
O
CM sr
to r^
< <:
z z
z z
Alkylation
Carbonylation
Of Of
«— in
0 NO
ro
IN-
in CM
CM O
S" m
0 NO
o
R 1
CO 0
«— o
o o
0 0
in CO
1^ NO
in o
o r—
CM O
CO
0 ON
gj ~
in «—
O CO
OJ *O
Z 2E
:z z
g
Catalytic reform
Condensation
OC Of
fC CM
CM in
ON
in
NO 5
tO CM
to in
o
in
CM CO
to
0 0
o o
CD CD
•O r—
8 8
CM NO
CD
O
r-^ in
in
in
in -O
si- ro
in to
< <
^ z
< <
z ^
c
o
Dehydrogenation
Dehydrohalogenat
o£ a:
gsf
to
IN. CM
to
3 |
8 §
O CM
CM
S 3
CM
>- O
O O
0 O
in co
^ >o
CM O
O O
*- CM
O
^ 3
in
CO Kl
in *o
O CO
z i
i i
Esterif ication
Generic reaction
Of Of
to
o
NO
sf
in
u
o
o
o
0
o
CD
CD
3
IN-
3
CO
CM
to
-
Z
Halogenation
of
vf fO
sr to
o o
r-.
8 3
in
NO
in CM
o to
f-- O
O CO
«— ON
CO CD
«— O
O 0
O O
1C R
o
° ^
§o
in
CO
•
N- CO
«— O
CO ON
$ z
z z
c
Hydrodimerizatio
Hydrogenation
Of Of
CO CM
in si-
NO
o in
CM in
s
«- CM
IN- in
O
o «-
CM N.
in
o
CO 0
*- o
o CD
0 0
1C 3
3 5
CM CO
CD
CM
ON -»
R
in »4-
£ 5
z z
z z
c
o
Hydroformlyation
Hydrohalogeneati
Of Of
CM CM
o in
0 0
0
Sf CM
O CO
IO
CM CM
o in
o
st fM
° "
CO CO
o o
o in
NO Is—
O NO
CD NO
o
ON ON
sr
._
ON CO
in sj-
z z
«c <
Hydrolysis
Nitration
Of Of
r- -a
O 0
0
CM
| o
st- CM
«— to
O
O CD
to
CD CO
0 «-
0 0
ON NO
CO
to
~ fe
sj-
o
0
o r-*-
in
in
OJ «-
«- CO
OJ *r-
z z
z z
Oxidation
Oxyacetylation
of rf
to o
o
o o
CM
n
in o
in
o
•*
a
CO CO
CD O
IN- S-
0
C
fO CD
CD NC
tO K
*~
NO
sf JO
Z Z
2 z
Oxyhalogenation
Sulfonation
Of Of
I
c •
O '"N
• w M
II D
U
0 Q
E1 O
CO ^
cu
a
x s a
•— nj o
< > z
2-26
-------
was reviewed and the following reaction types were selected
for generating Type B model streams:
(10) Hydroformylation
(11) Hydrogenation
(12) Hydrohalogenation
(13) Hydrolysis
(14) Nitration
(15) Oxidation
(16) Oxyacetylation
(17) Oxyhalogenation
(18) Sulfonation
(1) Alkylation
(2) Carbonylation
(3) Catalytic reformation
(4) Condensation
(5) Dehydrogenation
(6) Dehydrohalogenation
(7) Esterification
(8) Halogenation
(9) Hydrodimerization
The data base entries were sorted by reaction type, and the
model development approach outlined in Section 2.2.2.1 was
used to generate the model streams for these 18 reactor
processes. A separate generic model stream was developed to
represent reaction types that were not included in the 18
specified reaction types listed above. The model stream for
this "unspecified reaction" type was developed using data from
all reactor vent streams and the approach outlined in
Section 2.2.2.1.
2.2.3 Model Characteristics
Twelve example model process vent streams are presented
in Table 2-6 to illustrate the potential emission reductions
and cost impacts involved with controlling the process vent
streams. These model streams were selected to illustrate a
range of impacts as well as a range of production processes
and control technologies.
2.3 ENVIRONMENTAL AND ENERGY IMPACTS OF CONTROLLING EMISSIONS
FROM PROCESS VENTS
This section summarizes the environmental and energy
impacts associated with combustion control of process vent
streams. The environmental impacts of this control technique
include air and water pollution, waste disposal, pollution
prevention, and energy use. Combustion control devices such
as thermal incinerators and flares destroy organic compounds
through thermal oxidation and, therefore, reduce potential HAP
and VOC air emissions from process vents. The applicability
2-27
-------
CO
H
p^
EH
CO
EH
H
>
CO
CO
H
U
O
ft
ni
w
Q
i
H
ni
ft
§
w
vo
1
CM
W
J-3
CO
rfj
EH
j
Control
Method
_. o
8*1
8*5
* §
W g
^
tLj"
,, c o
•£ Q) W
« C 3"
W O -P
u pa
— '
0)
-P ^
*n
|5 SJ
O *"
H
Pn
!>i ^^
i) -P j_j
c'o ^
^ rt§
U
1) f
§ I
> EH
u
£
n
'a
T
_J S r-
Q (TJ C
13 ?j g
>2 -p £
co 5s
C C
O O
-H -H
Thermal
Incinerat
Thermal
Incinerat
s a
^* O
• •
o •*
in CM
o o
in o
oo in
CTI
0
in H
CO
< <
S-i
•H
rd
Formaldehyde Vi.
Oxidation
Air Oxidation
rH CM
C C
O O
-H -rH
Thermal
Incinerat
Thermal
Incinerat
55 55
CO CTi
CTI in
CM r-
0 0
O 0
CO ***!'
H CO
rH VO
rH
O
CTi CTI
r*» co
^
< <
ll
•H 5H
-H
0)
•H -H
Phthalic Anhydr
Oxidation
Terephthalic Ac
Oxidation
ro ••*
^_i
O in
Incinerat
& Scrubbe
X
in
c*^
CM
CO
o^
CO
VO
CO
Q
Distillation NV
in
c
o
-H
Thermal
Incinerat
53
O
CM
CO
in
CM
o
CM
co
Q
Distillation V
vo
2-?fl
Flare
Flare
55 55
O
O
CM CTi
CM
CO
CM
co in
•* in
t~» o
o
in en
o co
^
Q Q
Ethylbenzene
Forma Idehyde
r~ co
, C
O
•H
Thermal
Incinerat
55
O
0
o
CTi
rH
H
O
O
CTi
ex;
c
(d o
Adiponitrile Vi
Hydrodimerizati
CTI
Flare
x
CM
O
^
VO
f~
ro
O
o
CM
VO
&
rH
t>1
-P
Q)
O
C
O
S id
•H
Ethylene Glycol
Ether Acetate V
Ester if icat ion
o
rH
SH
O >H
Incinerat
& Scrubbe
x
r^
vo
CO
CO
CO
CM
CO
VO
CO
ex;
Halogenation
H
rH
Flare
a
0
o
^
in
CM
in
o
0
o
CM
H
O
ex;
Condensation
CM
H
ream.
-P
tfl
-P
c
>
!H
O
-P
O
(Ti
IVJ
0)
II
II
ex;
*^
6
(d
Q)
|-j
-P
£
0)
^
o
-p
(0
rH
'"j
Tj
tn
•H
Q
Q
• *»
g
cii
w
p?
1 1
r oxidation venl
; Y = Yes.
-H O
•< 55
I! II
i< S
Cd rQ
-------
and costs of these technologies are discussed in Volume IB of
this document.
Combustion devices are universally applicable for the
control of VOC's and HAP's in all process vent streams.
Combustion control is much less dependent on process and vent
stream conditions than other control techniques. Incinerators
and flares are the only demonstrated VOC combustion controls
that are applicable to all reactor processes. Flares,
however, can only be used on vent streams containing
nonhalogenated organic compounds due to corrosion to the
equipment and scrubber control requirements for halogenated
acid gases. Both incinerator and flare costs and efficiency
determinations require a limited amount of vent stream data
(i.e., volumetric flow rate, VOC emission rate, heat content,
and corrosion properties). The analysis of incinerator and
flare combustion control techniques yields conservative
estimates of energy, economic, and environmental impacts.24
The primary air pollution impacts are discussed in
Section 2.3.1. Section 2.3.2 discusses the secondary
.pollution impacts. Other impacts such as water pollution,
solid waste, pollution prevention, and energy impacts are
discussed in Section 2.3.3.
2.3.1 Primary Air Pollution Impacts
The achievable reduction of HAP and VOC emissions through
combustion control devices is based on the characteristics of
the process vent emission stream such as flow rate, heat
content, and HAP concentration, and the combustion device
design. Table 2-7 presents HAP and VOC baseline emissions and
emissions reductions achievable with combustion control for
the example process vent streams.
2.3.2 Secondary Air Pollution Impacts
This section evaluates the on-site secondary emissions
associated with combustion control. The secondary air
pollution impacts associated with combustion control include
emissions of nitrogen oxides (NOX) and CO. These secondary
emissions for the example process vent streams are presented
in Table 2-8. Secondary air impacts result from the
2-29
-------
I~
tn ^*1
£ u"
"° >
1 ^
S
Q) ""*
«
C '-»
O £4
W |U »s.
r\ W ft
& -g a
3 1 u
^ 11 tn
w s
0
1 rH "T~"
g ° £
O s Ic °
(H "3 ° >
^i M ^
Q cri H S
W II _p "-'
^ EH 1 "H
^ *& Ij ri3 ^^
2 > 1 w >,
" W -H ^"
H pj |l »*s ^N
O 11 "H
0 ^ I ^ S
el I tr
^ o 1 >,
ft c ^
1 tn
> -H g
i ii PJ ^^*^
CM 1 W
pt| I] Q) ^-»
t-1 ^
PQ "
S Q) ft
s ^
Cn
E
, e ^
1r
S^
II w 2
in
en o co o
c^ co in co
co c^ en
f» in
^i-
en o en o
C^ rH in CO
t^. t^ en
CM in
CM
CM O > CM
rH in H CM
H H rH
CM ^l-
CM in r- CM
rH in rH CM
rH rH
t^.
rH O in O
en ^ r~ H
in co H
c^ vo
VO
H O VO O
en r^ t-» o
r-» co H
CM VO
H CM CO ^
VO
O rH CO
rH 0 CO rH
• • • •
CO CM CO O
H
VO
VO CO H CO
r~ CM co H
vo vo co o
en
VD f»
^ t~- CM
CM ^l1 VO O
O CM O O
CM O O O
en
vo >
r~ r~ CM
en CM vo o
VO rH O O
H O O O
en
co co
H CM CO H
O CM CO O
H H
en
VO VO CO CO
^- CO CO rH
co vo co o
in vo f» co
**
CM
rH CO *3*
VO t-» CO O
H H H
H
rH
CO
o co ^r o
Cn rH rH
^
in
vo
o "* co
vo in co o
CO CO CO O
CM O O O
in
c-
co o vo
«* t-> o o
CO CM CO O
rH O O O
CO
CO
in CM •*
co c- en o
rH rH H
H
f»-
CO
CM VO O CO
CM co in o
Cn rH rH
0 H CM
°* H H H
2-30
-------
to
CO
W
CO
EH
W
CO
CO
W
U
0*
,
W
iJ
OH
§
M
g
Pn
CO
EH
U
rtj
H
!x
&
Q
55
O
U
w
CO
•
CO
1
eg
w
s
EH
!>, M
Jj 4J *~
•H § >
-H | ^
.p M f
° "5 s
«
U)
I-p —
0) >
W t T^"
o.»j -3
Q) *->
»
4J
S5r
c, d) ^1
/*\ 3 K*"
^ 53 ^
(tj _i j
t^ "H (o
3 2,
0)
«
IQ
C
0
•H —
01 k
CQ ^i
-H — .
E CT>
W S
^— '
X
0
(fl
c
o
"w "^
to ^
•H j_
w 5,
o
u
, 2 ^
^ 2 "i
a^g
o
co o
I * *
ID- O O 0
O «3 O
O
ID
r>-
•
CO O O O CM
CO O O O
CM O O
CO ID CO
CO
O O O O
O O O O
• • • •
O O O O CO
o
CO
H
0
yj oj
H 0
co vo o
o vo a\ o
O ID CO ^ O
ID r>3
CO VO
Oi
"* CM
CO ID O
o to o
O T 'T 00 O
O CO O CO O
r>j t~i
•H CM CO ^ ID
O O O O
O O O O
• • • •
*sj* O O
co r^ cr>
H
o
•
CO O O O VO O O
H H
OOOOO O
OOOOO O
o o o o o o o
0
o
o
CO
•S
•
H CO CO H
VO ^J< VO ID i* I
CO H H VO H CO H
H O O CM O H O
O O O ID O i-H O
— . . . . . _ „,
H H O O O ID O
O O O ^S* O O O
o o o o o o o
vo r- co a\ ° H N
H.H
II
'
•
to
CM
-------
combustion of organic HAP's and VOC's and from the combustion
of auxiliary fuel used for the combustion device and for
generating electricity. Fuel combustion for generating
electricity is assumed to occur off site and, therefore, the
impacts are not included in this discussion.
2.3.3 Other Impacts
Other impacts to be considered from the use of combustion
devices can include water pollution, solid waste, pollution
prevention, and energy.
2.3.3.1 Water Pollution Impacts. Control of VOC and HAP
emissions using thermal oxidation does not result'in any
significant increase in wastewater discharge; that is, no
water effluents are generated by the thermal oxidizer.
The use of an incinerator/scrubber system for control of
halogenated VOC vent streams results in increased water
consumption. In this type of control system, water is used to
remove the acid gas contained in the thermal oxidizer outlet
stream. The amount of wastewater generated is equal to the
amount of water needed by the scrubber to absorb the acid gas
leaving the incinerator.25 The water pollution impacts from
scrubber use are presented in Table 2-8. The increase in
total plant wastewater load would be relatively small and
would not affect plant wastewater treatment or sewer capacity.
However, the absorbed acid gas may cause the water leaving the
scrubber to have a low pH. This acidic effluent could lower
the pH of the total plant effluent if it is released into the
plant wastewater system. Some process units may recover the
acidic scrubber effluent for reuse or resale.
The water effluent guidelines for individual States may
require that industrial sources maintain the pH of water
effluent within specified limits. To meet these guidelines,
the water used as a scrubbing agent may need to be neutralized
prior to discharge to the plant wastewater system. The
scrubber effluent can be neutralized by adding caustic (NaOH)
to the scrubbing water. The amount of caustic needed depends
on the amount of acid gas in the waste gas from the combustion
device. For example, approximately 1.09 kg (2.4 Ib) of
2-32
-------
caustic are needed to neutralize 1 kg (2.2 Ib) of hydrochloric
acid.26
The salt formed in the neutralization step must be purged
from the system and properly eliminated. The methods of
disposal include direct wastewater discharge or salt recovery.
Salt recovery is only justified for large vent streams
containing a high percentage of halogens. In developing the
cost impacts presented in Section 2.4, the cost of caustic
needed for neutralization was not included. Further, the
costs associated with the disposal of the salt were not judged
to be significant in comparison to the control costs, and
therefore, were not included in the projected impacts.26
The use of scrubbers to remove hydrochloric acid from the
incinerator exhaust gas also has the potential to result in
small increases in the quantities of organic compounds
released into plant wastewater. However, only small amounts
of organic compounds are released into the scrubber
wastewater, and the flow of wastewater from the scrubber is
small in comparison to total plant wastewater, especially in
installations where there are multiple chemical processing
units using a central wastewater treatment process unit.
Therefore, the increase in the release of organic compounds in
plant wastewater is not likely to be significant.27
2.3.3.2 Solid Waste Disposal Impacts. There are no
significant solid wastes generated as a result of control by
thermal oxidation. A small amount of solid waste for disposal
could result if catalytic oxidation, instead of thermal
oxidation, were used by a process unit to achieve an
equivalent degree of VOC or HAP control. The solid waste
would consist of spent catalyst.
2.3.3.3 Pollution Prevention. Pollution prevention
involves the reduction of releases to the environment through
internal process improvements. For example, emissions from
process vents can be reduced by installation of a product
recovery device such as a condenser.
In the HON analysis, impacts for controlling process
vents were estimated based on combustion control devices that
2-33
-------
destroy the emissions instead of recover material from the
emission stream. Destruction of organic compounds in a
combustion device such as a steam-generating unit may replace
natural gas as a fuel, thus reducing the use of conventional
fuels and yielding some net benefit to the environment.
However, this type of control would not be considered a
pollution prevention measure.
2.3.3.4 Energy Impacts. The use of combustion devices
to control HAP process vents can result in a net energy
savings in some cases, while in other instances a net fuel
usage results. The use of an existing boiler or process
heater for control of energy-rich streams usually results in a
net fuel savings. An extremely low energy value for a process
vent stream may severely compromise the heat production rate,
however. The use of an incinerator results in a net energy
usage if supplemental fuel is needed to support combustion or
to promote flame stability. Flares can also require
supplemental fuel for flame stability if the heat content of
the vent stream is very low. An increase in the combustion
efficiency from a State-mandated control to the maximum
achievable control technology (MACT) level may also increase
the auxiliary fuel requirements. Impacts from increased fuel
demand for the example process vent streams are presented in
Table 2-8.
2.4 COST IMPACTS OF CONTROLLING EMISSIONS FROM PROCESS VENTS
The costs of controlling air emissions of organic HAP's
from process vents depend on the vent stream characteristics
and the type of combustion device used. The methodology for
calculating costs of combustion controls is discussed in
Volume IB of this document. For this analysis, it was assumed
that each production process would be equipped with a
dedicated combustion device. Some cost savings could be
achieved at larger process units if a common combustion device
was used to control multiple production process vent streams.
Table 2-9 summarizes the cost impacts of controlling each
2-34
-------
CO
2
W
CO
EH
S3
W
>EL PROCESS
u
O
a
W
i-3
S
§
^
W
0
En
CO
H
EH
irf
jg
H
EH
CO
W 1
EH
CO
O
0
^
1
u
II
ti« II
3
* II
en
CM
H
ij
9
EH
"
8*
§£
-p > S
M -H •>
O 4-1 f.
u o 2
CU *
W "-
c
OC Emissioi
Reductions
'Mq VOC/vr)
^> ^^
*-*^
co
co
cu
jj 0) — «•
jjj I> CJi
O {/)•
cu ^
tl 1
til
w
CJ
o w
c c
5< -H -rH H
S to 4J r"
K (Q U >•
i r^ D^
r-j 2 *"j ^J
* w w — '
o ^
EH
.p
W
O ^^
(0 1 ^"1
o 2v>
EH g «—
C
<
"^ to cu
*O U "S
o i S
S4*** »3
w &
^
n
in o
• •
">
•^ CO
~, H
0 H
m H o co
> co in co H
co r^ en
r~- in
co
in CM
• •
VO t^ O Cn rH O O
H O O CO VO O
"''i* CM ^d* CD en
en co
H
VO CO H
^* t^* (N CO
o^ c> ^^ £5 ^o vo n
t^ H in co
p» t^ en
CM in
CO
co
O O O O CO O O
H O O O O O
CO O O O «* H
H O CM O VO VO
o in co in ^
H rH VO
H CM
H CM CO •<* in VO >
O CO O O O
o vo r^- r- o
o in vo r-» o
> CM in t-»
CO 0
CO rH
VO <3-
CO CM
H rH CO ^J1
o vo r- co o
H H H
H
O CO O O O
O CM CM O O
O t^» 'S' 'S1 O
r^ co t~> co
CO fO
CO rH
VO rH
CO CO
O O CO •* O
Cn rH H
o o o o o
o o o o o
o co co o in
vo in in en in
^ vo »* o ^j*
„ O H CM
W W H rH rH
c
o
•H
4->
D
TS
CU
PH
C
o
•H
CO
-H
HAP/VOC Em
rH
(0
4->
O
EH
I
'I*
^_^
t .
>•
o
EH
II
g
^•~-
^
/eness
K'
4->
O
0)
4-1
4-J ^
W V
K
Wy,
C
o £E
u ^«
m
2-35
-------
example model stream. These costs are based on the vent
stream flow rate, heat content, HAP concentration, and size of
the combustion device.
2-36
-------
2.5 REFERENCES
2,
3.
4.
5.
6.
7.
8.
9.
10,
11,
12
13
U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Reactor Processes in
Synthetic Organic Chemical Manufacturing Industry -
Background Information for Proposed Standard.
EPA-450/3-90-016a. Research Triangle Park, NC. June
1990. p. 3-1.
Ref. 1, p. 3-2.
U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Air Oxidation Processes
in Synthetic Organic Chemical Manufacturing Industry -
Background Information for Proposed Standard.
EPA-450/3-82-001a. Research Triangle Park, NC. October
1983. p. 3-1.
Ref. 1, p. 3-5.
Ref. 1, p. 3-11.
U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Distillation Operations
in Synthetic Organic Chemical Manufacturing Industry -
Background Information for Proposed Standard.
EPA-450/3-83-005a. Research Triangle Park, NC. December
1983. pp. 3-8 to 3-11.
Ref. 6, pp. 3-17 to 3-22.
Ref. 3, p. 3-1.
Ref. 3, p. 3-19.
Memorandum from Olsen, T., K., Scott and B. Ferrero,
Radian Corporation, to Evans, L., EPA/CPB. September 17,
1991. Development of model emission source
characteristics for SOCMI process vents.
Memorandum from Olsen, T., Radian Corporation, to
Evans, L., EPA/CPB. September 1, 1991. Data used in HON
process vents data base development.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 2: Acrylonitrile Manufacture.
EPA-450/3-73/006b. Research Triangle Park, NC.
February 1975. pp. AN-10 to AN-15.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 3: Ethylene Dichloride Manufacture by
Oxychlorination. EPA-450/3-73/006c. Research Triangle
Park, NC. November 1974. pp. ED-10 to ED-16.
2-37
-------
14,
15,
16.
17.
18.
19.
20.
21.
22.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 4: Formaldehyde Manufacture with the
Silver Catalyst Process. EPA-450/3-73-006d. Research
Triangle Park, NC. March 1975. pp FS-11 to FS-16.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 5: Formaldehyde Manufacture with the
Mixed Oxide Catalyst Process. EPA-450/3-73-006e.
Research Triangle Park, NC. March 1975. pp. FM-11 to
FM-12.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 6: Ethylene Oxide Manufacture by Direct
Oxidation of Ethylene. EPA-450/3-73-006f. Research
Triangle Park, NC. June 1975. pp. EO-10 to EO-14.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 7: Phthalic Anhydride Manufacture from
Ortho-Xylene. EPA-450/3-73-006g. Research Triangle
Park, NC. July 1975. pp. PA-10 to PA-13.
U.S. Environmental Protection Agency. Engineering and
Cost Study of Air Pollution Control for the Petrochemical
Industry, Volume 8: Vinyl Chloride Manufacture by the
Balanced Process. EPA-450/3-73-006h. Research Triangle
Park, NC. July 1975. pp. VCM-13 to VCM-20.
U.S. Environmental Protection Agency. Survey Reports on
Atmospheric Emissions from the Petroleum Industry,
Volume'I. EPA-450/3-73-005a. Research Triangle Park,
NC. January 1974. Tables ACD-III, AC-III, HAC-III,
ACA-III, ACE-III, ANA-III, AA-III, AN-III, AL-III.
U.S. Environmental Protection Agency. Survey Reports on
Atmospheric Emissions from the Petroleum Industry,
Volume II. EPA-450/3-73-005b. Research. Triangle Park,
NC. April 1974. Tables CD-Ill, CO-III, DT-III, EL-III,
EDC-III, FS-III, G-III, HCN-III, TDI-III.
U.S. Environmental Protection Agency. Survey Reports on
Atmospheric Emissions from the Petroleum Industry,
Volume III- EPA-450/3-73-005C. Research Triangle Park,
NC. April 1974. Tables MA-III, NYL-III, N6, 6-III,
OA-III, PM-III, HP-Ill, LP-III.
U.S. Environmental Protection Agency. Survey Reports on
Atmospheric Emissions from the Petroleum Industry,
Volume IV. EPA-450/3-73-005d. Research Triangle Park,
NC. April 1974. Tables PS-III, PV-III, ST-III, SB-Ill,
VI-III, VAC-III, VC-III.
2-38
-------
23. IT Enviroscience. Emissions Control Options for the
Synthetic Organic Chemicals Manufacturing Industry.
Prepared for U.S. Environmental Protection Agency, Office
of Air Quality Planning and Standards. EPA Contract
Number 68-02-2577. July 1980.
Ref. 1, p. 4-1.
Memorandum from Olsen, T. and K. Scott, Radian
Corporation, to HON project file. February 6, 1992.
Estimating secondary impacts for HON transfer and process
vents.
26. Ref. 1, p. 7-7.
27. Ref. 1, p. 7-8.
24,
25.
2-39
-------
-------
3.0 TRANSFER LOADING OPERATIONS
Transfer loading operations within the SOCMI have the
potential to generate emissions containing VOC's, many of
which are HAP's. The emissions occur through vapor
displacement of VOC's.while a transfer vehicle is being
loaded. Transfer loading emissions consist of (1) the
evaporation of residual products from previous loads,
(2) emissions from the transfer vehicle as the chemical
product is loaded, and (3) vapors lost from the transfer
piping during the entire loading cycle.
Emissions from transfer loading can be reduced by a
variety of techniques, including the installation of a vapor
balance system or a vapor collection system that would route
the emissions to a control device. Numerous control
techniques are suitable in specific cases, but the most
universally .applicable technologies for controlling emissions
from transfer loading operations are based on combustion
(e.g., flaring, incineration, or incineration with scrubbing).
A more detailed discussion of control techniques applicable to
transfer operations is presented in Volume IB of this
document.
To estimate the impacts of various control options for
transfer loading operations, model transfer racks were
developed. Transfer racks have been characterized by the
number of chemicals transferred at the rack and the volume of
the chemicals transferred. Using these model transfer racks,
the environmental (e.g., emissions and reductions), energy,
and cost impacts of the control options were assessed.
This chapter presents the assessment of environmental,
energy, and cost impacts. To aid in the overall understanding
3-1
-------
of this assessment, Section 3.1 includes a brief description
of transfer loading operations and the potential emission
sources at transfer racks. The development of model transfer
racks is presented in Section 3.2. Section 3.3 discusses the
environmental and energy impacts of controlling emissions from
transfer racks and Section 3.4 presents the cost impacts of
the control technologies.
3.1 EMISSION SOURCE DESCRIPTION
Liquid products and co-products of SOCMI processes are
transferred from storage vessels to transport vehicles through
loading racks. Equipment such as pumps, meters, piping, and
grounding are used to fill tank trucks and tank cars through
loading arms on the racks.
The two principal sources of emissions from transfer
operations are loading losses and fugitive emissions.
Evaporative losses from loading can be released directly to
the atmosphere from the tank truck or tank car during loading,
while fugitive emissions can be released through the hatches
and other openings of tank trucks and tank cars that are not
vapor tight.
Loading losses are the primary source of HAP emissions
from tank truck and tank car loading operations. Loading
losses occur as organic compound vapors in "liquid-empty"
transport vessels are displaced to the atmosphere as the
liquid chemical is loaded into the vessels.
Loading losses occur by three mechanisms: displacement
of vapors that are transferred into the vehicle via the vapor
balance system as the previous product was unloaded;
displacement of vapors formed in the empty tank by evaporation
of residual products from previous loads; and vapor
displacement and volatilization as a result of turbulence and
vapor/liquid contact during loading of the new product. A
smaller amount of organic compound may be lost to the
atmosphere in the form of vapor that remains in the transfer
system after transfer loading operations have been completed.
The total amount of organic compound losses from loading
operations is, therefore, a function of the physical and
3-2
-------
chemical characteristics of the previous product, the method
of unloading the previous product, the type of operations used
to transport the empty carrier to a loading terminal, the
method of loading the new product, and the physical and
chemical characteristics of the new product.1
The two principal methods of loading are splash loading
and submerged loading. In splash loading, the fill pipe
dispensing the chemical is lowered only partway into the
transport vessel (i.e., barge, tank car, or tank truck).
Significant turbulence and vapor/liquid contact occur during
splash loading, potentially resulting in a high degree of
vapor generation and loss. if the turbulence is great enough,
liquid droplets can be entrained in the vented vapor.1 The
actual quantity of organic compound lost is dependent on the
physical properties of the chemical transferred, the transfer
rate, and the degree of turbulence in the transport vessel.
A second method of loading is submerged loading, of which
there are two types—the submerged fill pipe method and the
bottom-loading method. In the submerged fill pipe method, the
fill pipe extends almost to the bottom of the transport
vessel. In the bottom-loading method, a permanent fill pipe
is attached to the bottom of the transport vessel. Figure 3-1
presents a bottom-loading tank truck with a vapor collection
system. During both types of submerged loading the fill pipe
opening is below the liquid surface level for most of the
loading operation, thus minimizing liquid turbulence and
resulting in much lower vapor generation than occurs with
splash loading. Some of the advantages of bottom loading are
improved safety, faster loading, and reduced labor costs.1
The historical use of any given transport vessel can be
as important in determining loading losses as the method of
loading. if the transport vessel previously carried a
nonvolatile liquid such as propylene glycol, the vapor space
in the transport vessel will contain little or no VOC's. This
would also be true for transport vessels that are cleaned
before use, which is a common practice in the chemical
industry. On the other hand, if the transport vessel has just
3-3
-------
O
C
(0
•P
O
M-J
O
(0
M
tP
C
•H
H
I
CO
0}
Cr>
•H
3-4
-------
carried a volatile product, such as ethyl chloride, and has
not been cleaned, the vapor space in the transport vessel will
contain VOC's, which will be expelled with newly generated
vapors during loading.
Some cargo carriers are designated to transport only one
product; this practice is called "dedicated service."
Dedicated transport vessels return to a loading terminal with
a vapor space that is fully or partially saturated with the
organic compound from the previous load. Transport vessels
may also be "switch loaded" with various products, so that a
nonvolatile product may be loaded to expel the vapors
remaining from a previous load of a volatile product. Switch
loading can be used when handling petroleum products, but it
is less frequently used in the chemical industry where product
purity is a principal concern. The potential for these
situations vary with the type of transport vessel, the cargo
carrier, the chemicals being transported, the geographic
location, and the season of the year.1
3.2 MODEL EMISSION SOURCE DEVELOPMENT
As part of the development of the model transfer racks,
Section 114 questionnaires requesting data on the loading
practices for HAP's or HAP/VOC mixtures were sent to nine
corporations. The collected data were used to evaluate
impacts and to develop model transfer racks for tank trucks
and tank cars. The information was used to relate rack size
(number of loading arms) to the number and volume of chemicals
loaded.
3.2.1 Data Gathering
Under authority of Section 114 of the Clean Air Act
(CAA), nine corporations were asked to complete questionnaires
on transfer operations associated with the SOCMI. The list of
corporations receiving Section 114 questionnaires was
developed to maximize the number and variety of product
processes sampled.
The questionnaire requested information on facility
transfer operations, but focussed on the chemical processes
that have organic HAP's as products or co-products. The
3-5
-------
requested information included chemical production capacity,
quantity of material loaded as a function of chemical
production capacity, types and loading techniques of transport
vessels, size and operation of the transfer loading racks, and
vapor collection and emission controls. Representatives of
each corporation were contacted (as necessary) to clarify
information in the Section 114 responses and to obtain
information on the relationship between the number of loading
racks dedicated to each chemical process and the amount of
material transferred by pipeline. The Section 114
questionnaire responses and the information collected from the
telephone conversations with representatives of the
corporations surveyed were compiled into a data base.
Table 3-1 contains the transfer rack characteristics
contained in the data base that were used to develop model
transfer racks. Most information in the data base was
compiled directly from the Section 114 responses. However,
some information required interpretation before it could be
used in the development of model transfer racks or the
assessment of impacts. For example, some Section 114
responses cited several types of transport vessels loaded by a
single transfer rack. In these instances, only a portion of
the volume of chemicals loaded at the transfer rack was
designated for a specific transport vessel type.2 This
procedure resulted in several entries for a single transfer
rack at a single facility location. Another example of data
manipulation before evaluation involved Section 114 responses
reporting the use of two (or more) transfer racks for a single
chemical. In these cases, the quantity of the chemical
transferred at each transfer rack was proportioned to the size
of each rack.
Several fleets of transport vessels were reported for
some individual facilities in the Section 114 responses. For
these facilities, the total number of transport vessels for
each facility was calculated as the sum of the number of
transport vessels in each fleet. The number of transport
3-6
-------
TABLE 3-1. TRANSFER RACK CHARACTERISTICS3
(1) Model rack name
(2) Number of chemical production units
(3) Number of chemicals transferred
(4) Throughput
(5) Transport vessel type
(6) Loading technique
(7) Number of transport vessels in fleet
(8) Number of transport vessels in fleet equipped
with a vapor collection system
(9) Maximum chemical transfer rate per arm
alnformation used to determine model transfer racks.
3-7
-------
vessels with vapor collection systems was similarly
established for facilities reporting multiple fleets of
transport vessels.
3.2.2 Model Development
Model transfer racks for tank trucks and tank cars were
developed using the information gathered from the Section 114
questionnaires. In addition, allocation factors were
developed to estimate the volume of chemical products
transferred in each type of transport vessel. This section
describes the development of the model transfer racks,
characterization of the displaced vapor, and generation of the
allocation factors.
3.2.2.1 Model Transfer Rack Development. The
Section 114 information was reviewed to relate rack size
(number of loading arms) to the number of chemicals
transferred at a given transfer rack and the total volume of
chemicals transferred. Two transfer rack characteristics were
selected for this evaluation because a single loading
parameter did not provide a clear relationship. Models were
developed by reviewing the rack sizes (number of loading arms)
relative to the number of chemicals loaded and the volume of
chemical loaded. Table 3-2 presents the model transfer racks
for tank car operations, and Table 3-3 presents the model
transfer racks for tank truck operations.
The development of the model transfer racks was an
iterative process using the Section 114 information as a guide
for evaluating.validity. The number of racks and loading arms
predicted by the model transfer racks were compared to the
number of racks and loading arms reported in the Section 114
responses. Based on the results of this comparison, the
throughput ranges for each model transfer rack were modified
several times for each range of number of chemicals
transferred. The transfer rack characteristics shown in Table
3-2 and Table 3-3 represent the final model transfer racks for
each transport vessel type.
3-8
-------
TABLE 3-2. MODEL TANK CAR TRANSFER RACKS
Number of
Materials
1-3
4 - .9
10 - 22
> 23
Throughput. (TP) Range
(MMgal/yr)
0 < TP < 10
10 < TP < 40
40 < TP < 80a
0 < TP <10
10 < TP < 20
20 < TP < 30
30 < TP < 60b
0 < TP < 3
3 < TP < 80°
0 < TP < 10
10 < TP < 20d
Number of
Arms
3
8
16
3
6
10
16
3
10
4
9
aFor throughputs above the maximum value, add an additional
3-arm rack per 10 MMgal.
bFor throughputs above the maximum value, add an additional
3-arm rack per 10 MMgal.
GFor throughputs above the maximum value, add an additional
3-arm rack per 3 MMgal.
dFor throughputs above the maximum value, add an additional
4-arm rack per 10 MMgal.
3-9
-------
TABLE 3-3. MODEL TANK TRUCK TRANSFER RACKS
Number of
Materials
1-4
5-12
13 - 20
> 21
Throughput (TP) Range
(MMgal/yr)
0 < TP < 3
3 < TP < 12
12 < TP <. 70a
0 < TP <. 3 . 5
3.5 < TP < 7.5
7.5 < TP < 21
21 < TP < 54b
0 < TP < 30C
0 < TP < 12
12 < TP < 24d
Number
of
Arms
1
2
4
1
2
4
6
1
3
4
additional
aFor throughputs above the maximum value, add an
1-arm rack per 3 MMgal.
bFor throughputs above the maximum value, add an
1-arm rack per 3.5 MMgal.
°For throughputs above the maximum value, add an
1-arm rack per 15 MMgal.
dFor throughputs above the maximum value, add an additional
4-arm rack per 12 MMgal.
additional
additional
3-10
-------
The number of model racks predicted using these tables
was slightly lower for facilities producing a large number of
chemicals with a small total chemical throughput (e.g., a
specialty chemical manufacturing complex). In addition, for
operations manufacturing few chemicals yet loading great
volumes of chemicals, the number of predicted transfer racks
was slightly larger than the reported number of transfer
racks. However, for the SOCMI as a whole, the predicted
number of transfer racks and loading arms based on the model
transfer racks realistically represents actual loading
operations and practices. Further improvement of the model
transfer racks would require extensive facility-specific data
and would not be expected to improve the representativeness of
the existing model transfer racks.
3.2.2.2 Loading Rack Vent Stream Characteristics.
Estimating the cost of controlling emissions from transfer
operations requires vent stream data that represent the
characteristics of the displaced vapor. The stream
characteristics included in the analysis of impacts are flow
rate, annual hours of operation, molecular weight, and heat
content.
The vent stream flow rate from the model loading rack to
the control device is a function of the transfer rate and the
number of loading arms. An average transfer rate per loading
arm was estimated using the Section 114 information. This
rate was then applied to each model transfer rack to determine
the vent stream flow rate for the entire model transfer rack
(i.e., the average flow rate per arm times the number of
arms). The average transfer rate per arm for model tank truck
transfer racks was found to be 170 gal/min and the average
transfer rate for model tank car transfer racks was calculated
to be 190 gal/min.2 The average transfer rates were then used
to estimate the volume of vapor displaced during transfer
operations. When calculating these average transfer rates per
arm, it was not considered that some transfer racks employ
vapor balancing to reduce releases to the atmosphere.
Therefore, the average transfer rates determined in this
3-11
-------
manner are expected to yield conservatively high estimates of
emissions from transfer operations when applied to the entire
industry.
The annual hours for transfer, which were necessary to
estimate control device operating costs, were generated by
applying the transfer rate to the total quantity of material
loaded for the model transfer rack.2 This procedure yields a
representative average operating time for the vent stream.
The control device operating time may be different, however,
depending on the frequency of transfer operations. For
example, if an incinerator is used to control emissions from
transfer operations, the time required for controlled start-
up or stand-by operation will be greater than merely the time
required to fill the transport vessel.
Another important characteristic to consider for the
proper design of an emission control device is an average
molecular weight for the organic compound(s) handled. The
average molecular weight for the chemicals transferred at each
transfer rack was calculated as follows:
Molecular _
Weight :
HAP
Emission
lb-mole\
yr /
HAP
. * Molecular!
Ib \
T .HAP /
."^-Emission \
lb-mole\
yr /
Moxecu±ar ,•
Weight i \lb-mol/
where:
HAP Emission (lb-mole/yr)= Throughput * Emission
Factor.
These average molecular weights provided representative
inputs for the molecular weight ranges for each organic
compound included in the Section 114 responses.
Vent stream heat content was estimated in a manner
similar to that for the molecular weight. The heat content
was estimated using the heat of combustion of each organic
compound, the emissions of each organic compound, the
operating time of the transfer rack, and the total volume
transferred through the selected model transfer rack. The
3-12
-------
natural gas requirement for the combustion control devices was
a function of the vent stream heat content. This requirement,
therefore, was not biased high or low, because the heat of
combustion of many chemicals was used in developing the
average heat content for any given transfer rack.
3.2.2.3 Allocation Factors. In response to the
Section 114 questionnaire, chemical manufacturing facilities
reported on the percentage of their chemical production
capacity that is transferred by tank truck, tank car, marine
vessel, and pipeline. Based on this information, a throughput
allocation factor was developed for each transport vessel type
for each facility.2 This factor is used to predict the volume
of a chemical transferred at a SOCMI facility as a function of
the production capacity for each chemical process.
The development of allocation factors was segmented
according to facilities with marine terminals and facilities
without marine terminals. The information contained in the
Section 114 responses showed a measurable difference in the
relative quantities of chemical transferred to tank trucks and
tank cars, depending on the presence of marine vessel
transfer. For each model transfer rack, the volume of
chemicals transferred to each type of transport vessel (i.e.,
tank truck and tank car) was calculated and totalled. The
total volume of all of these chemicals produced at the
facility was also calculated. The allocation factor for the
facility was determined by dividing the total volume
transferred by the total volume produced. The allocation
factors presented in Table 3-4 represent the average of the
allocation factors for the facilities that responded to the
Section 114 questionnaire.
Two scenarios were developed to represent the information
contained in the Section 114 responses. Scenario One
represents facilities that load only tank trucks and tank
cars. Scenario Two represents facilities that load tank
trucks, tank cars, and marine vessels. Under Scenario One,
the tank truck allocation factor is 2.78 percent of the
production volume and the tank car allocation factor is
3-13
-------
TABLE 3-4. THROUGHPUT ALLOCATION FACTORS
Scenario
Tank Truck
Tank Car
Onea
Twob
2.78
4.28
4.22
9.76
aScenario One corresponds to facilities that load tank
trucks and tank cars.
^Scenario Two corresponds to facilities that load tank
trucks, tank cars, and marine vessels.
3-14
-------
4.22 percent of the production volume. Under Scenario Two,
the allocation factors are approximately two times those for
Scenario One. The tank truck allocation factor is
4.28 percent of the production capacity and the tank car
allocation factor is 9.76 percent of the production capacity.
These data lend some'insight into the types and locations of
facilities that responded to the Section 114 questionnaire.
Facilities that have marine terminals ship a greater volume of
chemicals than those without marine transfer operations. This
result indicates that major chemical complexes producing bulk
intermediate organic chemicals have been sited to maximize the
flexibility of shipping products (by truck, rail, or barge).
The analysis of allocation factors included the
production volume of chemicals provided for captive use in the
total production volume for a facility. If considering the
source of an allocation factor for a single, specific
chemical, these average allocation factors would tend to
estimate low quantities transferred. However, these factors
are considered representative when taken for the entire
industry. They become even more useful when applied to the
industry production volumes in a manner consistent with this
development.
3.3 ENVIRONMENTAL AND ENERGY IMPACTS OF CONTROLLING EMISSIONS
FROM TRANSFER LOADING OPERATIONS
This section presents impacts for controlling emissions
from several example model transfer racks by applying a
combustion device. Impacts were developed for example
facilities contained in the data base for SOCMI. The
aggregate transfer quantity for all chemicals manufactured at
the facility was used to assign a model transfer loading rack.
Impacts are shown for tank car and tank truck loading; impacts
for tank car and tank truck loading are similar because the
emissions and costs are related to the amount transferred and
the use of combustion control, and not to conversion of the
loading rack. The impacts are presented in terms of
emissions, emission reductions (and increases of secondary
pollutant emissions), water pollution, solid waste generation,
3-15
-------
and energy consumption. In addition, the costs of .the control
options for the transfer loading racks are presented.
3.3.1 Primary Air Pollution Impacts
Primary air emission impacts that result from control of
emissions from transfer operations are reductions of the VOC's
and organic HAP's. These reductions are directly related to
the quantity of chemicals transferred and to the individual
chemicals transferred. For the HON analysis, impacts of
controlling emissions from transfer operations are based on
combustion in a flare or incinerator. As shown in Table 3-5,
most of the transfer racks for which impacts were developed
were controlled using a flare.
The estimated emissions of organic HAP's from the
transfer racks are shown in Table 3-6. The associated
emission reductions for organic HAP's are also shown in the
table, and represent the 98 percent reduction achievable
through combustion control. These reductions range from
3.56 x 10~6 Mg/yr to 19.3 Mg/yr. As anticipated, the
emissions reductions from transfer operations x-epresents the
smallest potential reductions for any of the five major
emission source types in the SOCMI.
3.3.2 Secondary Air Pollution Impacts
This section evaluates the on-site secondary
emissions associated with combustion control. The secondary
air pollution impacts associated with combustion control
include emissions of NOX and CO. These secondary emissions
for the example transfer racks are presented in Table 3-7.
Secondary air impacts result from the combustion of organic
HAP's and VOC's and from the combustion of auxiliary fuel used
for the combustion device and for generating electricity.
Fuel combustion for generating electricity is assumed to occur
off site and, therefore, the impacts are not included in this
discussion.
The combustion control used to reduce emissions from
transfer operations results in a relatively small increase of
secondary pollutant emissions, ranging up to 0.13 Mg/yr of NOX
and up to 0.025 Mg/yr of CO. These impacts are more the
3-16
-------
tfl
w
o
f§
«
fa
is
cei
p!
W
Q
O
-1
1— 1
cu
X
w
in
i
CO
H
m
EH
O
O
•H
£>
JH g "C;
(fl O fj) HH
p^4 »A 'pj
Qj (d /i\ S
icf! **""''
CU
4J —
g 0 r-l
5 f?{ ^»
| 43 \
X3 rd
O &i
S §
EH ---
O rH
rd
J-i O
(U -H
g 0)
536
ol a
is
S 0
rH Q
T5 ^
g^
H M
pf
\.
o
(d
0)
C
•H
O
G
-rH
rH
rH Q)
rd 45
rd 0) SH
rH 43 O
fa EH W
^r ^j
o
rH
CM O
0 0
H
H
CO in
CO CO
O VO
0 0
'
H H
CO CO
tl (l
(d (d
o o
C C
cd fd
B EH
rH CM
0)
rH
rd
rH
fa
J2;
O
in
CM
O
o
CO
o
rH
CO
rH
rd
o
C
EH
CO
\
O
4-1
fd
0)
G
-H
O
fi
-rH
rH 0)
rd 45
g 45
rH 3
0) M
43 O
EH W
X
CTi
^*
H
in
•<*
•*
CO
rH
rd
0
G
EH
•*
Q)
fd
H
fa
g
fd
rH
0)
C
•H
O
G
•H
rH
rH Q)
rd 45
g 45
S 3
0) rH
43 O
EH W
X
CTl
CO
CO
CO
CO
CO
•*
rH
rH
rd
o
fd
EH
vo
d> 0)
rH S
rd rd
rH rH
fa fa
is is
Ol CM
in <*
VO CO
CT> in
CM VO
rH rH
H CM
CO CO
tt tl
fd rd
O O
«*• ±1*
G G
rd fd
EH EH
r~ co
ti tj
0 O
-P 4-1
rd rd
(i ti
0) (1)
C C
•H -H
O O
G G
•rH -rH
rH H
rd rd
g g
a) a)
43 43
EH EH
63 65'
CM
in r>
H CM
CO
VO CM
in CM
<*• CM
••* in
vo o
H rH
rl H
rd rd
O O
KJ 1J
G G
rd (d
EH EH
o
rH
0)
rd
rH
fa
55
o
H
CM
O
CO
*3*
in
o
0
H
rH
O
^
-P
G
fd
EH
rH
rH
^
O
4_)
fd
JLj
0)
G
•H
O
G
•H
rH
rH 0)
fd 45
Q) M
43 0
EH W
X
0
O
H
in
CO
VO
o
H
CO
O
rH
-p
G
rd
EH
CM
rH
0) 0)
rd rd
rH rH
fa fa
65 !S
CO' VD
o vo
t-» in
r- co
CM VO
CM in
CM ^,CM
rH CM
o a
? ?
M rH
-P -P
•s -g
fd ra
EH EH
co <#
rH H
3-17
-------
w
^£4
u
3
w
Cn
K
EH
H °
5 H
o 5
^ H
H ^
& °
§ ^
X
w
.
m
i
co
rl
EH
Q)
O
•H
^
Q)
Q
i-H
0
^_j
-p
c
o
u
o
a
fird
0) u
0 -P
rH fd
fd -P
a w
Q) g
en J-i £3 *y^
2 & ™ §
> >
ci 1^ W
| ft >
"** r^ irt
^ P (0
j£j r^ ©
EH ^
O rH
fd
M U
(!) -H
e a)
M M
X Q)
rS O XI
° « §
S S
O
•p
(0
^_(
0)
fi
•H
U
c
•H
J_l
rH Q)
fd Xt
£l .Q
H
en
H
H
VO
in
r-
'
•
o\
H
O
^
c
rd
EH
in
H
(1)
03
rH
fe
•2;
CN]
en
m
CO
r-
o
H
"*
O
d
-P
c
H
11
^
o
S
(d
3-18
-------
OH
O
Cn
S3
O
C_i
~
U CO
ID to
Q «
W U
(X H
Q
Q 0
<;
H
c< P-l
3
S
• H
vo
1
CO
W
CQ
EH
• in
c
o
•»-J
rn
D
-o
CJ)
04
•H
in
in
•H
rH
O
S_j
^H
-P
c
o
CJ
j^
0)
4J
S(
rj
o
^
a^
£
^^
^H
^
\
<
w
Cn
—
^^
M
>i
CJ
0
>
2
A
,_^
M
p^T
^
Cn
S-
_
J_)
^
^*
Co
,_^
j_i
>i
Qj
HH
K
Cn
"-
k
* Q)
o ja
(d g
"1
VO
i
o
H
*
VO
in
CO
VO
1
o
H
*
vo
in
CO
CO
1
o
1
t — 1
*
0
o
r-
CO
1
o
rH
*
0
CO
r-
VO
i
0
H
*
CO
vo
CO
vo
0
rH
*
CO
vo
CO
H
•*^CNCN H VO'*'S<'*HHH
1 1 1 1 1 1 1 1 1 1 1 1
oooo o ooooooo
HHrHH^H^Q^^rHrHHrHHHrH^
•x***0'?*^^1^*******^
O3O3COCnCOCNVO03COrHin'51'
CO**OH vo cNcocnr^oHin
vorHcocn cn oivooicnoiojco
•^•^CNCN rH VD^^f^rHrHrH
1 1 1 1 1 1 1 1 1 1 1 1
OOOO O OOOOOOO
rHrHHH{y1Hr^(y)lf)rHHHHHHrHvo
•K***^*^^^*******^
O3CNCOCnr)COr^VOlfl'*CNVOCNCOrHin':1<
CO^OH vo cococn>OH.in
voHcocn cn ojvoo3cno303co
invO^COCNOlCNrHrHCOinvOincOCOCNO]
1 1 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1
ooooooooooooooooo
*****************
cncncncoHvooioinocn'*coinO'^co'~lkDir>cn«*o3O3ojino]':tl
-tf^HCO CO CO-*OCnHrH>
VOrHCOCn Cn O3VOCOCnCNO3CO
«,^.^5^'^ILOVOt^cO
ojco^invoococnHHHHHHHHH
3-19
-------
Ql
O
53
C EMISSION REDUCTIO
TRANSFER RACKSa
TINUED)
O M 53
> W O
Q U
g o —
^
O . 1 1 |
^ g
5
?s
vo
1
en
PH
l_^
CQ
<3
^
IQ
C
o
•H
•P
O
TS
0)
q
o
•H
l/l
u
•H
g
w
Ions After Control
Emiss:
SUOT!
to
U]
•H
g
W
0)
q
•r-
"oi
en
x^
o-
O
CT>
* '
^
J^
r>^
PT
^4
W
&^
g_
_< ^
n j^j /t\
QJ f\ f\
g PH P
JZJ
CM CO
CM •
• O^
^* t~H
CM CO
CM
• C\
•* H
CM H
1 1
O O
H H
* *
H in
vo en
co co
CM H
I I
o o
H H
•K *
H IT)
VO CTl
* •
CO CO
o t-
co •
• en
^* t — 1
0 t-
/T\
" _J
*^* t~H
o\ o
H CM
device .
rH
O
-P
o
o
T!
0)
-P
(8
O
•H
•3
0)
Ti
frt
U
(0
"^
y
o
(0
J_l
Q)
q
a
-P
X)
o
id
w
(0
3-20
-------
(C
CO
u
K
.
W
1
EH
Q
W
1
W
0
EH
U
^4
s
H
O
H
EH
g
O
H
>H
1
Q
§
U
CO
.
CO
W
iJ
CO
<
EH
O ^
LJ Ti ^
QJ ft^
fl* 3
O to
f> w ^^
u w IT
•H g
W
10
c ^-«
O to
X--H >
O W ~^>
alg
•J7
y .
CO f^
rr* (^
XI
H g
a) 3
•o a
o
s
CM
O
rH
0 *
' CO
f^
CM
rH H
1 1
0 0
H H
* *
CO CO
Cn rH
cn co
^o <&
i i
o o
* *
H r-
O CO
cn r-
in in
i i
0 O
* *
in CM
CM ^O
CM CO
H CM
o in CM co -*
HO O O O
rH H rH rH
*
o o o • * ooooo* oo* oo*
o ^r co ^o co
rH ** t^ H «*
\o ...
. H CM VO CO
H
rHCMCMCOCOCMCMt7l7iCMCM'--|r>JogCM
OOOOOOO1-.1— OOOOOO
HHrHHHrHH~j;^rHrHrHrHrHH
CM H rH
in*******^^. ******
covocor>rH'aco
vocMcncot^t<~cMQj_|coincoincMcn
^rHCOrHHCMH^^CMinHr^'i'H
in'*COCOCMCMCOCOVOVOCO(0-^
-------
CO
g
CO
EH
1
a
W
i-q
0»
W
0
CO Q"
EH W
LLUTION IMPAC
(CONTINU
O
05
H
s
s
i
o
u
W
CO
r-
CO
TABLE
c
Water
nsumptio
[gal/yr)
o
u
w c
to o Tr
°43>
*/* & 2
i § 4^
3 u S
•PCg
(ti O S.
55 0
CO
Emissions
(Mg/yr)
c -->
0 M
X-H >
o w --•
a w tr
•H a
M '
In
OJ 0)
X)
H g
Q) 3
T> 55
O
O
H
*
in
0
CO
H
O
*
in
in
i
o
iH
*
O
in
CO
o
H
*
o
CO
co'
H
O
CO
o
H
*
O
co
CM
CM
1
O
i-H
*
in
CM
CM
o
H
*
vo
CO
VO
0
CM
•*
0)
c
Refere
g
o
$
3-22
-------
result of the combustion of supplemental gas than the
combustion of the organic HAP's in the vent stream.
3.3.3 Other Impacts
3.3.3.1 Water Pollution Impacts. Water pollution
impacts may result from the control of emissions from transfer
operations if the vented gases are halogenated compounds. In
these cases, the control system would include a scrubber to
remove acid gases generated in the combustion process. The
amount of wastewater generated is equal to the amount of water
needed by the scrubber to absorb the acid gas leaving the
4
incinerator. The resulting wastewater stream would require
treatment before being discharged to the nearest surface water
or to the local publicly-owned treatment works (POTW). The
amount of water generated from control of emissions from
transfer operations is small, as shown in Table 3-7. No
adverse impacts on water pollution are anticipated, however,
as a result of controlling emissions from transfer operations.
3.3.3.2 Solid Waste Disposal Impacts. No solid waste
impacts from the control of emissions from transfer loading
operations are anticipated.
3.3.3.3 Pollution Prevention. Pollution prevention
involves, the reduction of releases to the environment through
internal process improvements. For example, the reduction of
emissions from transfer operations by using a pressure
transfer instead of a transfer to a transport vessel vented to
the atmosphere would qualify as a pollution prevention
measure. This type of transfer is common in the industry for
compounds with higher vapor pressures, which warrant transfer
under pressure. Combustion (i.e., flare incinerator,
incinerator with scrubber) is the control measure evaluated in
the HON analysis for reducing emissions from transfer
operations. This type of control would not be considered a
pollution prevention measure. However, if the organic
compounds combusted replace natural gas as a fuel, there would
be some net benefit to the environment through the
substitution of fuels.
3-23
-------
3.3.3.4 Energy Impacts. The energy impacts from the
control of emissions from transfer operations will include the
cost of fuel to supplement the firing of the incinerator used
to combust the organic HAP's in the vent stream. The flue
requirements are dependent on the flow rate and the heat
4 *
content of the VOC stream. This can result in a net cost of
energy to the process unit. For example, Table 3-7 shows the
gas requirements for the example transfer loading racks.
There is a net gas requirement for each rack ranging from less
than 1 MMBtu/hr to over 2,800 MMBtu/hr. If, however, the flow
rate and heat content of the emission stream from transfer
operations is sufficient to support combustion and the waste
heat can be used elsewhere in the process unit, there will be
a net benefit to the control. Site-specific conditions and
needs would dictate the value of heat recovery to the process
unit. Such site-specific conditions were not evaluated as
part of this impacts analysis.
3.4 COST IMPACTS OF CONTROLLING EMISSIONS FROM TRANSFER
LOADING OPERATIONS
The costs of the control systems for transfer operations
depend on the transfer rate for a given loading rack for the
entire facility and the combustion control device, selected.
The costs in this analysis are based on the use of a single
transfer loading rack for the entire facility eind a single
control device servicing that rack. Chapter 3 of Volume IB of
this document contains the detailed information on development
of costs for the vent collection system and control device.
The costs of the control systems for the model transfer
racks are given in Table 3-8. No product recovery is assumed
with the use of combustion control devices. The total annual
*
costs of the control systems for model transfer racks range
from $6,870 to $84/400. With the small emission reductions
anticipated for controlling emissions from transfer
operations, the cost effectiveness ratios for these transfer
loading racks are relatively high, ranging from $l,990/Mg HAP
to $2.93 billion/Mg HAP.
3-24
-------
TABLE 3-8. ANNUAL CONTROL COST ESTIMATES
FOR EXAMPLE MODEL TRANSFER RACKSa
- Model Total
Rack Annual
-Number Cost ($/yr)
1 9,
2 63
3 9,
4 66
5 10
6 84
7 22
8 25
9 39
10 28
11 6,
12 63
13 10
.14 15
15 63
16 16
17 14
18 74
19 67
20 38
630
,800
650
,000
,600
,400
,600
,900
,000
,100
870
,800
,100
,600
,800
,800
,300
,300
,200
,400
Total HAP/VOC
Emission
Reduction
(Mg/yr)
3
6
1
3
9
9
2
6
2
9
2
2
8
.56
.32
.42
.08
.19
3
.63
1
6
5
.34
.32
.96
.72
.08
.11
.55
4
4
*
*
*
*
*
10
10
10
10
10
-6
-4
-4
-2
-2
.39
*
10
-1
.67
.39
.65
*
*
*
*
*
*
*
10
10
10
-6
-4
-4
10-4 .
10-1
10-1
10
-1
.46
.22
19.
3
Cost
Effectiveness
($/Mg)b
2.
1.
6.
2.
1.
2.
2.
1.
6.
4.
2.
1.
3.
1.
3.
7.
1.
1.
1.
1.
70
01
81
14
16
49
35
55
11
98
93
01
42
61
06
97
68
67
59
99
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
109
108
10?
106
105
104
104
104
103
103
109
108
10?
10?
105
104
104
104
104
103
aEach transfer rack has a dedicated control device.
bCost Effectiveness ($/Mg) = Total Annual Cost ($/yr) -s-
Total HAP/VOC Emission Reduction (Mg/yr).
3-25
-------
3.5 REFERENCES
1.
U.S. Environmental Protection Agency.
Pollution Emission Factors, Volume I:
and Area Sources Location. EPA/AP-42,
Park, NC. September 1985. p. 4.4-3.
Compilation of Air
Stationary Point
Research Triangle
2.
3.
4.
Memorandum from Ocamb, D. and T. Olsen, Radian
Corporation, to Markwordt, D., EPA/CPB. October 7, 1991.
Revised HON transfer model loading racks development.
Memorandum from Olsen, T., Radian Corporation to
Stelling, J., Radian Corporation. August 21, 1991.
Loading techniques used in the SOCMI.
Memorandum from Olsen, T. and K. Scott, Radian
Corporation, to HON project file. February 3, 1992.
Estimating secondary impacts for HON transfer and process
vents.
3-26
-------
4.0 STORAGE TANKS
This section discusses potential emission sources, models
used to represent storage tank farms, and impacts of control
technologies for storage tanks in the SOCMI. Section 4.1
presents a brief description of the sources of storage tank
emissions. Section 4.2 discusses the development of model
storage tank farms. The environmental and energy impacts of
controlling emissions from storage tanks are presented in
Section 4.3. The cost impacts of the control technologies are
given in Section 4.4
4.1 EMISSION SOURCE DESCRIPTION
There are four primary types of vessels commonly used for
storage in the chemical manufacturing industry:
• Fixed roof tanks;
• Internal floating roof tanks;
• External floating roof tanks; and
• Horizontal tanks.
Fixed roof and floating roof tanks are cylindrical vessels
with axes perpendicular to the ground. The axes of horizontal
tanks, also cylindrical in shape, are parallel to the ground.
Because none of these vessels is air-tight, air may enter
the tank and become laden with organic HAP vapor. The
displacement of this vapor-laden air, which occurs during tank
filling or as a result of diurnal temperature and barometric
pressure variations, is the source of emissions. Eguations
for calculating emissions from storage tanks are given in the
EPA report "Compilation of Air Pollutant Emission Factors,
Volume I: Stationary Point and Area Sources."
4-1
-------
4.1.1 Fixed Roof Tanks
Of currently used tank designs, the fixed roof tank is
the least expensive to construct and is generally considered
to be minimum acceptable equipment for the storage of organic
HAP's. A typical fixed roof tank, shown in Figure 4-1,
consists of a cylindrical steel shell with a cone- or dome-
shaped roof that is permanently affixed to the tank shell. A
conservation vent, which is a type of pressure- and vacuum-
relief valve, is commonly installed on fixed roof tanks to
contain minor changes in vapor volume. Because* this valve
prevents the release of vapors during only small pressure
differentials (±0.2 kPa), emission reduction can be as low as
5 percent depending on the vapor pressure of the stored
liquid.
The major types of emissions from fixed roof tanks are
breathing and working losses. Breathing loss is the expulsion
of vapor from a tank vapor space because of diurnal changes in
temperature and barometric pressure. Emissions; can occur in
the absence of any change in the liquid level in the tank.
Working losses occur during filling when the liquid level
in the tank rises. Vapors are expelled from the tank when the
pressure inside the tank exceeds the relief pressure as a
result of filling.
4.1.2 Floating Roof Tanks
A storage vessel can have an internal floating roof or an
external floating roof. An internal floating roof tank has
both a permanently affixed roof and a roof inside the tank
that floats on the surface of the stored liquid. The floating
roof rises and falls with the liquid level in the tank.
Circulation vents and an open vent at the top of the fixed
roof are usually provided to minimize the possibility of
organic HAP vapors accumulating in concentrations approaching
the flammable range.
Loss of organic HAP vapors from internal floating roof
tanks can occur in the following four ways:
(1) Through the annular rim space around the perimeter of
the floating roof (rim or seal losses);
4-2
-------
Pressure Vacuum Valve or
Conservation Vent
Gauge Hatch
Manway
External Level Gauge
Manway
Submerged Fill
or Discharge
' /'
Figure 4-1. Typical fixed roof storage tank.
4-3
-------
(2) Through the openings in the deck required for various
types of fittings (fitting losses);
(3) Through the nonwelded seams formed when joining
sections of the deck material (deck seam losses); and
(4) Through evaporation of liquid left on the tank wall
following withdrawal of liquid from the tank
(withdrawal loss).
As wind flows over the exterior of an internal floating
roof tank, air flows into the enclosed space between the fixed
and floating roofs through some of the shell vents and out of
the enclosed space through others. Any organic HAP that has
evaporated from the exposed liquid surface and that has not
been contained by the floating deck will be swept out of the
enclosed space.
All internal floating roofs have a closure device to seal
the gap between the tank wall and the perimeter of the
floating roof. A primary seal may be liquid mounted or vapor
mounted. Whereas liquid-mounted seals rest on the surface of
the stored liquid, there is a vapor space between the stored
liquid and a vapor-mounted seal. Secondary seals may be used
to provide some additional evaporative loss control over that
achieved by the primary seal. The secondary seal would be
mounted to an extended vertical rim plate located above the
primary seal.
The numerous fittings that penetrate or are attached to
an internal floating roof include access hatches, column
wells, roof legs, sample pipes, ladder wells, vacuum breakers,
and automatic gauge float wells. Fitting losses, which occur
at these openings, can be controlled with gasketing and
sealing techniques or by the substitution of fittings that are
designed to leak less.
Deck seam losses are inherent in several floating roof
types. Any roof constructed of sheets or panels fastened by
mechanical fasteners (e.g., bolts) is expected to have deck
seam losses. Deck seam losses are considered to be a function
of the length of the seams and not the type of mechanical
fastener or the position of the deck relative to the liquid
4-4
-------
surface. The control for deck seam losses is achieved by
selection of a roof type with vapor-tight deck seams. Welded
deck seams are vapor tight and not a source of emissions.
External floating roof tanks do not have permanently
affixed roofs. A floating roof is the only barrier between
the stored liquid and the atmosphere. The types of emissions
from external floating roof tanks are seal losses, withdrawal
losses, and fitting losses. Roof fittings for external
floating roof tanks include access hatches, guide-pole wells,
gauge float wells, vacuum breakers, roof drains, roof legs,
and rim vents. External floating roof tanks do not have deck
seam losses because they are constructed of welded steel.
4.1.3 Horizontal Tanks
Emissions from horizontal tanks are similar to those from
vertical fixed roof tanks—breathing losses and working
losses. According to earlier EPA studies, emission equations
for fixed roof tanks may also be used to provide reasonable
approximations of emissions from horizontal tanks. Although
the correlations are not directly applicable, the fundamental
concepts do apply.
4.2 MODEL EMISSION SOURCE DEVELOPMENT
To assess the national impacts of regulating storage
tanks at SOCMI facilities, the impacts on each individual
SOCMI facility in the country could be examined. However,
because detailed information needed for such.an assessment was
not available for each facility, baseline emissions and
impacts for each facility were estimated by developing model
storage tank farms to approximate the storage practices found
in SOCMI facilities.
As described in Section 4.1, there are four types of
storage tanks being used by the SOCMI. Some types of tanks
are control techniques for other types of tanks. For example,
because internal floating roof tanks have lower emissions than
fixed roof tanks, an internal floating roof may be installed
as a control for a fixed roof tank. Many States require this
type of control as reasonably available control technology
4-5
-------
(RACT) to satisfy the requirements of their State
Implementation Plans (SIPs).
For use in characterizing the storage practices in the
SOCMI, model tanks must be representative not only of the
sizes of tanks that are common in the industry but also of the
types of tanks. For this reason it was necessary to review
State regulations to determine the current (baseline) level of
control required in existing facilities.
The following sections describe the procedures used to
select model tank sizes and to determine model tank types
based on baseline control levels.
4.2.1 Data Gathering
The primary sources of information used in the
development of model storage tanks were EPA reports on organic
chemical manufacturing and storage of VOC's. Both the organic
chemical manufacturing (OCM) report and the volatile organic
liquid (VOL) storage BID contain data on tank population and
tank size distribution in the United States. Current level of
control was determined by reviewing existing State regulations
for storage operations. Additional information on State
regulations is presented in Volume 1A of this BID.
4.2.2 Model Development
There were three major activities associated with model
tank farm development. The first, tank sizing, involved
selection of typical tank capacities. The second step was to
develop for each tank size a throughput range and the number
of tanks in a tank farm. Third, a review of State regulations
determined current levels of control. The following sections
discuss these three activities and the eight model tank types
that were developed.
4.2.2.1 Selection of Model Tank Parameters. In
developing model tanks, the four major parameters were
capacity, diameter, number of turnovers, and number of tanks
per size category in the model tank farms. The first step was
to choose the tank sizes. Based on a review of the OCM
report,4 which summarized data on the 1977 storage tank
population, 10,000 gal was chosen to be the smallest tank
4-6
-------
size. The other tank sizes (see Table 4-1) selected had
capacities that are common in the industry. The 10,000-gal
model tanks are horizontal; all other model tanks are
vertical.
In industry practice, the ratio of tank height to tank
diameter (H/D) rarely exceeds unity. In general, dimensions
for the model tanks were determined from the averages of
values found in the OCM report and the VOL storage BID.6'7
For all model tank sizes except the 40,000-gal tanks, the
average values met the expected criteria of H/D less than or
equal to 1.0. For the 40,000-gal tanks, a slightly larger
diameter was chosen to achieve H/D less than or equal to 1.0.
The annual number of turnovers for a tank is related to
the tank size. The OCM report contains a curve relating the
two parameters.8 The numbers of turnovers shown in Table 4-1
were taken from the OCM curve.
To characterize the storage operations associated with a
production process, it was necessary to develop model tank
farms to represent the number of tanks that would be used to
store a single chemical. Although several tanks in a variety
of sizes may be used to store a single chemical, to simplify
the analysis it was assumed that a certain number of same-size
tanks would-be used. The OCM report contains information on
the distribution of tanks among the various sizes, but the
data in that report represent a broader segment of the SOCMI
than that covered by the HON. It was decided that a more
representative data set could be developed from data gathered
during earlier stages of the HON program under authority of
Section 114 of the CAA.
Responses to Section 114 questionnaires were reviewed
from hydrocarbon producers, hydrocarbon users,
chlorofluorocarbon producers, and ethylene dichloride
producers. Data on tank size and number of tanks were used to
develop a target tank size distribution for the SOCMI on a
national basis.
4-7
-------
CO
<3
fa
W
*y
S3
EH
W
Q
O
g
.
H
1
W
1-3
CQ
EH
^^
On
C_|
^_^
.p
£
3
o
ft
r- j
<»•— '
<-<
Orientatioi
t-
CO
V
P^
EH
H
O
CO
^^
(-{
•p
C
0)
rH
CO
H
0
H
rH
.
o
o
CN
O
CM
VO
rH
V
PH
V|
CO
*
CN
r-
cn
CO
H
t^
^
o
VD
rH
(0
O
•H
4J
(1)
O
O
O
O
O
O
*.
H
O
VO
p^
H
V
PH
EH
VI
O
*
CN
VO
H
CO
rH
H
t^
^}*
in
CO
rH
(0
o
•H
•p
!H
0)
0
o
o
«*.
o
o
o
«.
CN]
PH
EH
VI
O
•
VO
t-
H
CO
A
rH
H
r-
^*
in
CO
rH
(0
O
-H
•P
0}
0
o
o
o
0
o
hk
CN]
4-8
-------
An iterative process was used to develop the throughput
range associated with each tank size and the number of tanks
in each tank farm. Using the values in Table 4-1 for tank
size and number of turnovers and assuming a number of tanks
per tank farm, the following equation was used to calculate a
maximum throughput for each tank farm:
Throughput
(MMgal/yr)
Tank Size
(gal)
No. of
Tanks
No. of
urnover
per Year
The maximum throughput for each tank farm was set as the
minimum throughput for the next largest (in storage capacity)
tank farm, thus establishing throughput ranges that represent
a continuum of throughput values.
After developing the throughput ranges, the resulting
national tank distribution was determined.10 Facility-specific
data for storage throughput in the SOCMI was matched to the
HON throughput ranges, and the number of tanks that would be
assigned to each tank size was calculated and summed for all
production processes. This HON tank distribution was compared
to the target tank distribution. The number of tanks per tank
farm was varied and the throughput ranges were recalculated
until the HON tank distribution approached the target tank
distribution.
4.2.2.2 Existing Level of Control. State regulations
were reviewed for all 50 States to determine the current level
of control. Only 24 States have regulations for storage
tanks * The following discussion is based on the rules of
eight States—California, Illinois, Louisiana, Michigan, New
Jersey, Ohio, Pennsylvania, and Texas—which account for
approximately 70 percent of the SOCMI. These rules were
reviewed in detail for their requirements for storage vessels.
In all of these States, requirements are based on tank
size and chemical vapor pressure. If the State rule does not
specify requirements, a horizontal tank was assumed for
10,000-gal vessels and a vertical fixed roof tank was assumed
for the larger vessels. All chemicals having a vapor pressure
4-9
-------
greater than or equal to 14.7 psia were assigned to pressure
tanks that were assumed to have no emissions.
Most of the States surveyed require floating roofs
(internal or external) or vapor-recovery systems as tank size
and chemical vapor pressure increase. For the HON analysis,
it was assumed that internal floating roof tanks are used
because external floating roof tanks are not commonly used to
store organic HAP's. It was also assumed that vapor recovery
is achieved with a refrigerated condenser, because this
technology was judged to be the most economical add-on control
device for achieving the required reduction levels.
Percent vapor recovery is specified in State regulations
for California, Illinois, Michigan, New Jersey, Ohio, and
Pennsylvania. Louisiana requires a reduction equivalent to
that achieved by installing a floating roof. Biased on
emission reductions given in the VOL storage BID, refrigerated
condenser efficiency for units in Louisiana was assumed to be
93 percent.1
Instead of specifying a percent vapor recovery, Texas
requires that the partial pressure of the VOC in the emission
stream be reduced to 1.5 psia. To translate this requirement
into an equivalent percent reduction, emissions for each
organic HAP were calculated twice—once using actual HAP vapor
pressure and a second time using a vapor pressure of 1.5 psia.
Percent emission reduction was calculated using the following
equation:
Equivalent
Percent
Emission
Reduction
Baseline _ Emissions Using Vapor
Emissions
Pressure = 1.5 psia
Baseline Emissions
* 100
Based on this approach, the Texas rules are equivalent to the
following reductions for various vapor pressure ranges: For
chemicals having vapor pressures between 1.5 and 11.0 psia,
emissions must be reduced by 80 percent; for chemicals having
vapor pressures between 11.0 and 14.7 psia, emissions must be
reduced by 90 percent.
4-10
-------
The eight model tank types listed in Table 4-2 satisfy
all of the baseline control requirements of the eight major
SOCMI States. To simplify the analysis, it was assumed that
the refrigerated condensers in Louisiana, Michigan,
New Jersey, Ohio, and Texas, which are required to achieve 93
or 90 percent reduction, are actually achieving 95 percent
reduction.
Regulations were not reviewed in detail for each of the
50 States. Baseline control levels for tanks in Alabama,
Arkansas, Colorado, Connecticut, Hawaii, Kansas, Maryland,
Massachusetts, Nevada, New Hampshire, North Carolina,
Oklahoma, Oregon, Vermont, Virginia, and Wisconsin were
assigned based on a "typical" State regulation. This default
regulation reflects the level of control in Pennsylvania and
Michigan.
The remaining 26 States have no regulations on storage,
and it was assumed that facilities in those States would use
the minimum acceptable control—a horizontal tank or a
vertical fixed roof tank depending on throughput. Vapor
pressure was disregarded as a factor in determining baseline
control unless it was greater than 14.7 psia and therefore
required a pressure tank. Actual SOCMI facilities in these
26 States may be using internal floating roof tanks for
economic reasons—to achieve the cost savings associated with
reduced product loss. By assuming the lowest level of
baseline control, the HON analysis calculated a conservative
estimate of the impacts of controlling these sources.
Halogenated compounds, pesticides, and some glycol ethers
have been found to be incompatible with aluminum. Storage
of these chemicals in internal floating roof tanks, which
often have aluminum decks, can result in corrosion of the roof
and contamination of the product. For this analysis; it was
assumed that these chemicals are not stored in internal
floating roof tanks. Where a State regulation would allow
these chemicals to be stored in an internal floating roof tank
based on chemical vapor pressure, it was assumed that control
4-11
-------
TABLE 4-2. MODEL TANK TYPES
Model Tank
Type
Description
FXR_DEFAULT
FXR_RC_95
FXR_RC_85
FXR_RC_80
IFR_DEFAULT
IFR_2SEALS
HORIZONTAL
HORIZ RC 95
Basic fixed roof tank.
Fixed roof tank with a refrigerated condenser
having a removal efficiency of 95%.
Fixed roof tank with a refrigerated condenser
having a removal efficiency of 85%.
Fixed roof tank with a refrigerated condenser
having a removal efficiency of 80%.
Internal floating roof tank with a vapor-
mounted primary seal.
Internal floating roof tank with a vapor-
mounted primary seal and a secondary seal.
Basic horizontal tank.
Horizontal tank with a refrigerated condenser
having a removal efficiency of 95%.
4-12
-------
would actually be achieved through vapor recovery in a
refrigerated condenser.
4.2.3 Model Characteristics
Seventeen example model tank farms, shown in Table 4-3,
were selected to illustrate the potential emission reductions
and cost impacts that could result from controlling storage
vessels. These model tank farms were selected to provide a
manageable number of examples while still illustrating the
range of impacts. The example models also represent a range
of tank sizes, chemical vapor pressure, and control
technologies.
Two control technologies were evaluated for storage tanks
in the HON analysis: (1) tank improvements (i.e., installing
an internal floating roof inside a fixed roof tank or
upgrading an existing internal floating roof), and
(2) refrigerated condensers. The applicability and cost of
these technologies are discussed in Volume IB of this
document. Baseline control technologies and chemical
properties were the major factors in selecting the control
technology appropriate for a particular model tank.
4.3 ENVIRONMENTAL AND ENERGY IMPACTS OF CONTROLLING EMISSIONS
FROM STORAGE TANKS
This section evaluates the environmental and energy
impacts associated with controlling storage vessels. Analysis
of environmental impacts includes an evaluation of the
potential for air and water pollution, waste disposal, and
pollution prevention. Estimation of energy impacts is based
on the electricity requirement for refrigerated condenser
systems.
4.3.1 Primary Air Pollution Impacts
Within the SOCMI, most VOC's are stored as individual
chemicals, not as mixtures. Because in the HON analysis all
of the organic HAP's that are stored are also VOC's, HAP and
VOC emission impacts will be the same. These impacts are
presented in Table 4-4 for the 17 example model tank farms.
Emission reduction from baseline for these examples ranges
from 31 percent for upgrading an existing floating roof to
4-13
-------
CO
PM
j.
J2j
rtj
E"H
w
Q
O
g
P3
Hi
1
w
f>
1
"#
W
CQ
S
-Q
-o
O
tu
s:
"o
t-
*"*
o
u
g-S-g
O (H «*— "if
.- c 0 S
E QJ 03 *~
£i o
NJ rxl X *>** '' ' '*' " '
X X X
U_ U- U_
^-4 oj co ^r m
0
o
01
c
•r-
O
1 —
H-
• —
cd
CO
c
r-»
^
CO
co
CD
CD
CD
0
OJ
1^
1—
§
Lul
3
X
LU
CO
o
o
cn
c
•t—
o
q;
i —
(0
cn
c:
CO
CD
CO
CD
0
CD
CD
CD
OJ
CD
•— t
g
LU
3
e
'^
0
o
01
c
•r-
(d
O
r —
0)
cd 0)
w c
c o
cd
:n
^H CO
r^. co
CD CD
^
cn cn
CO
CD O
CD CD
0 0
CD O
CD OJ
CD
OJ
S <•>
H- H-
^ ZD
LU LU
o; or
X X
U_ LL.
co cn
0)
w
0)
"O
c
o
CJ
^
CO
t-H
*—*
OJ
•^
C3
C3
0
0
CM
CO
1—
§
LU
3
2
CD
i-H
Q) 0)
w w
CD 0)
"c ^
O 0
o o
31 OI
CO OJ
1— 1 •— 1
CD r--
to co
cn r*»
OJ OJ
CD CD
CD CD
0 0
CD 0
CD CD
OJ OJ
r-. ^-
ID CO
< J
1 i 1 Q£
3 ^
2 u-
<-* OJ
1— 1 •— 1
S
cn
TJ
rd
O
, ' —
0)
CO QJ
C "O
0) rd
c cn
O 0.
O ID
1C
1— 1 1— 1
LTi CO
.— 1 1— «
0
cn co
CO CO
"*
CD O
0 0
CD O
0 0
CD CD
O OJ
OJ
co r-^
i—
CO ID
J <
Q: LU
^ °i
X Qi
3 S
8
01
C
!T>
cd
o
•—
QJ
-Q
cn
Q.
^H
CO
CO
^r
CO
to
o
CD
CD
0
CD
OJ
^
S
LU
CD
J
LO
0
O
01
-t—
res
o
J—
-S
cn
CL
O.J
CO
,3-
r-.
CO
OJ
CD
CD
CD
O
CD
CD
O»J
t^
1
Lul
O
CO
o
cn
c
•i—
fd
O
r—
H-
03
(d
O)
CL
CO
CO
CO
CO
T-H
o
CD
o
CD
CD
CD
OJ
CO
1—
g
LU
I
•a
QJ
CD
QJ
cn
0
fd
31
4-14
-------
(d
S
3
fa
§
EH
j_q
fa"
Q
O
g
1
s
w
PH
O
fa
s
o
H
o
Q
W
PH
S
O
H
W
W
H
O
g
Q
§
O4
sa
.
^*
i
•**
H
(_3
CQ
EH
CO
*-*
»-t
O
-H
i j
-r^
Q)
c
O
•H
CO
CO
•H
w
t-H
O
i j
o
u
.
f\\
w
4J
<
0
•H
10
C/]
"C
t~
•^
CO
c
lissio
&
CU
c
•I-
"o;
CO
rd
CQ
"57
u
o
>^
Cn
S-
^^
^1
e?
•^
33
JJ,
g
^,
s
O
tr>
v—'
^^
J_l
Oi
Cn
j^
^-^
^^
S-l
u
o
s
*~*
^^
S-i
t>1
oT
ffi
S^
o
tl
1^1 ^
c ^
rd "§
p [ P
rH »
® e
'O t
O *"
g^
CO *3* CO in ^* H CM
or— CM •* CM r-» inint--oincovocMco
HCOCMVDC--rHrHCMCM'*COCMt-~rHVO'*VD
Hcovoint~- •<* coin COCM
rH rH
co 'd* co in ^i* rH CM
or^cM-^cM r« mmr--oincovocM.co
rHCOCMVOt~-rHHCMCM^llCOCMr-rHVO'a'*CO'*OrHCO'3'r*-CMCr\Ot^r~
O O O CO rH VO rH "VO OCMOOCTiCM^'^J'Cn
rH H rH H VO -*
inOHCMVO HHCO'*VO C7l
ocM'4CM'*'vt H VOCOCMt^VOCOVOCOO
HcOCMOOOr>-COCOCMVOa>CMVO'*OCOVO
CM co c^ in r^ •* ^* t"» CM cr> c~-
H H
CT\ •* CO CO O O
ovocor-» rH VOCOCM>VOCOVOCOO
rHCOCMOOOr^COCOCMVDOACMVO'a'OCOVO
cMcot^int^ ** ^3"t^ cMcnr^
r* . H
~.«^ ^._OrHcMco-sa
• rH
0) rd
o o
-H -H
> 4-1
0) C
T3 Q)
rH -H
o
t j £^
4J Q)
c >
O Q)
o co
tJ >H
cu o
| 1 (|_j
(d
O CO
•H 4->
T! o
Q) rd
-a a
"
CO T3
rd QJ
£! C
IJ "_Q
C g
rd 0
4J O
Q) CU
tnx!
rd 4->
o ><
4J rH
CO rH
(d
X! d
O 4J
rd O
W rd
rd
4-15
-------
97 percent for installing an internal floating roof in a fixed
roof tank.
4.3.2 Secondary Air Pollution Impacts
Only on-site combustion emissions were considered to be
secondary air pollution impacts of the HON. Because neither
of the control techniques evaluated for storage vessels is a
combustion device, combustion emissions (secondary air
emissions) from controlling storage tanks were not considered
as potential impacts.
In some cases, the product recovered from a condenser is
a mixture that is valuable only for its heating value when
burned in a combustion device. However, for the HON analysis,
it was assumed that the condensed liquid is a pure product
that can be returned to the storage tank and later sold or
used in the manufacturing process. Thus, there is no on-site
combustion associated with controlling storage tanks in the
HON.
4.3.3 Other Impacts
4.3.3.1 Water Pollution Impacts. There are two
potential sources of water pollution associated with tank
improvements and condensers. First, before an internal
floating roof can be installed or upgraded, the tank must be
emptied and cleaned. Some wastewater will be generated during
tank cleaning. Second, if water is used as the cooling fluid
in the condenser, there may be some concentration of metals
and solids due to the evaporation/condensation cycle. Neither
of these sources of water pollution will result in adverse
impacts on water quality.
4.3.3.2 Solid Waste Disposal Impacts. There are no
significant solid wastes generated as a result of control by
tank improvements or refrigerated condensers.
4.3.3.3 Pollution Prevention. As described in
Section 2.2.2 of Volume IB of this document, a condenser is a
product recovery device. If the recovered product is returned
to the storage vessel for sale or for use in the manufacturing
process, there is a reduction in the amount of raw materials
that must be used in the process. When floating roofs are
4-16
-------
used to reduce emissions, there is also a reduction in lost
product or raw material purchased. For example, controlling
Model Tank Farm No. 12 results in an emission reduction of
32 Mg/yr. Those 32 Mg are recovered by refrigerated
condensation and are returned to the tank farm. Thus, the
facility does not have to purchase additional material to
replace the 32 Mg that would have been emitted in the absence
of control.
If the recovered product is burned as fuel in a
combustion device, there is a net benefit to the environment
due to reduction in the usage of conventional fuels. However,
this approach would not be considered a pollution prevention
measure.
4.3.3.4 Energy Impacts. The only energy impact
associated with controlling storage vessels is an increase in
electricity use necessary for funning the refrigerated
condenser system. For the example models in Table 4-3, annual
electricity requirements range from 84 MW-hr to 609 MW-hr.
4.4 COST IMPACTS OF CONTROLLING EMISSIONS FROM STORAGE TANKS
The costs of controlling air emissions of organic HAP's
from storage vessels depend on the emission rate from the
vessel and the specific control device used. Some cost
savings could be achieved at larger facilities if controls
were centralized (e.g., all tanks in one tank farm vented to
the same refrigerated condenser). However, for the HON
analysis it was assumed that each individual tank would be
equipped with a dedicated control device. Chapter 3 of
Volume IB of this document describes the methodology for
calculating the costs of tank improvements.and refrigerated
condenser systems.
Table 4-5 summarizes the cost impacts of controlling each
of the 17 example models. The product recovery credit shown
is the value in dollars per year of the recovered product. It
is calculated by multiplying the emission reduction by the
unit price of the individual chemical. As a credit, this
value is subtracted from the "gross" annual cost to give a
"net" total annual cost. Where possible, chemical-specific
4-17
-------
in
I
•<*
CQ
•<
EH
tn
in
0)
C o
v S&
W-rH |
rj -P \
m %^^
4n
H
W
rj c G ~
. i^ O O »
"J § -H -H C"
ro P* m oJ ^
j 1 x*. « ***^ X.
o ciT w ° tn
-H 3
«!~
rH ""^ "t~.
(C 2 fjj r-l
o £• O —
c r i ^^
•^ > —
1 1 i^t
!=j
C ^3 O
o
rH
O
C
o
0)
EH
r-H
O
J_|
-P
C
o
u
llll
o
o
o
«.
f^
o
H
CO
O
H
in
H
CM
o
H
H
CO
CM
H
CO
rH
in
H
H
H
}_l
0)
(0
£
Q)
rrj
C
0
u
H
0
H
*.
"*3*
H
^3*
r-
CO
CO
en
P*
CM
in
o
5
<*
co
CO
CM
f^
m
LI
o
(0
r^
0)
TS
C
0
CJ
CM
0
O
m.
H
CO
CM
CM
^
H
CO
CM
O
en
in
in
t~.
o
en
t"^
ft j
o
0
&1
G
-P
O
rH
4H
rH
rH
(0
•P
W
C
H
CO
^— N
o
CM
X^ . v
•*
VO
H
cTT
VO
f*
H
rH
*""
vo
>
en
rH
f>
o
en
r*^
t \ i
o
o
Cn
G
-P
(0
o
rH
q_l
rH
rH
(0
-P
C
H
"
^^^^
0
inv
•h
H
^^
CM
r-
co
c\
in
en
CO
in
1
CM
«#,
CO
en
CO
CO
:*
en
CO
H
in
CO
vo
CO
CO
H
H
en
^
t^
CO
r^
H
H
H
in
rH
4H
0
o
M
o*
c
-p
o
rH
4-1
rH
H
(0
-P
C
H
-
^-^
O
CO
vo
*.
H
*—*
CM
r^
rH
vo"
VO
H
O
CO
CM
CO
CO
CM
rH
CO
[^
H
H
H
in
H
4-1
O
o
tn
c
•P
(0
o
rH
<4_l
H
rH
(0
-p
(0
H
CO
0
VO
«»
in
vo
in
in
CM
o
en
co
[S,
vo
H
r***
t**
in
r^
VO
O^
f"^
vo
H
JH
O
W
C
0)
c
o
u
„
0
CM
VD
w
CM
in
^
•<*
CO
VO
•*
vo
H
H
•*
CM
CO
CO
in
in
LI
0)
en
c
0)
£
o
u
H
rH
O
H
rH
*,
r-
H
0
CM
CO
^
CM
in
r^.
^f
in
o
co
H
r-
t^
in
vo
•*
in
in
LI
0)
tn
c
0)
'G
C
O
(D
CM
H
O
O
en
i^
in
in
r^
in
vo
CO
-*
en
• CO
CO
a\
r--
CM
r~
-------
(d
^
P^
^1
CM
i
o
g .
H
OH
§
rn
b
W S"
EH W
;T ESTIM;
(CONCLUI
UJ
0
U
-.
1
tH
r^
S
O
u
ANNUAL
in
w
J
CO
EH
^ \
Total
HAP/VOC
rH
o
o
0)
Product
o
£3
-p
-H
n
(d
EH
CQ
CO
CU
CO
Q) — .
O
tl 1
fl 1
w
CO
c
eductio:
(Mg/yr)
PH
_Q
^ "57
o <^
°^
•p ^
•H M
"d ^~i
0) \
&-I
u •—
-p
-H
•d
0)
*•» •C/5'
Q)
> "•"
o
o
0)
(0
chnology
0)
EH
r— I
o
o
o
t .
M
C2 f\\
p u)
^
o
co
H
-*
H
CO
H
CO
1
in
OJ
in
H
V.
rH
O
CO
VO
VO
O
o
tn
C
•H
-P
O
rH
(U
•d
(d
en
^
H
O
O
CN
OJ
vo
VO
^^
in
H
CO
H
^
S
^
^
o
CO
VO
VO
o
o
•H
-P
(d
o
rH
m
0)
•d
i
0)
o
o
0)
p^
1
^—x
-H
•d
CU
u
ecovery
o
V^
4J
CO
O
U
(Annual
CO
o
u
1
c
c
rtj
rH
(0
O
EH
CO
C
O
•rH
-p
0
•d
0)
c
o
•H
CO
CQ
•rH
g
U
O
\
<
a
rH
(d
-p
O
EH
•1-
IT
^
\
-P
cn
0
u
rH
(0
g
c
f^J
rH
4J
O
EH
II
1
CO
CO
(1)
^
••H
-P
O
0)
-------
price data were used. If no chemical-specific data were
available, an average price of $1.57/kg was used.
As shown in the table, total annual costs for the model
tank farms range from a savings of $280,000 to a net cost of
$556,000. Although both control technologies result in large
emission reductions and therefore, product recovery credits,
the higher capital cost of the refrigerated condenser systems
results in higher annual costs.
Cost effectiveness is calculated by dividing the annual
cost of control by the annual emission reduction. As shown in
Table 4-5, cost effectiveness for the example models ranges
from a savings of $l,630/Mg to a cost of $l07,000/Mg.
4-20
-------
4.5 REFERENCES
1. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Compilation of Air
Pollutant Emission Factors, Volume I: Stationary Point
and Area Sources. EPA/AP-42. Research Triangle Park,
NC. September 1985. pp.' 4.3-1 through 4.3-35.
2. Erikson, D.G. (IT Enviroscience). Storage and Handling.
In: Organic Chemical Manufacturing, Volume 3: Storage,
Fugitive, and Secondary Sources. Report 1. U.S.
Environmental Protection Agency, Research Triangle Park,
NC. EPA-450/3-80-025. December 1980. p. II-7.
3. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Control of Volatile
Organic Emissions from Manufacture of Synthesized
Pharmaceutical Products. EPA-450/2-78-029. Research
Triangle Park, NC. December 1978. p. 3-17.
4. Ref. 2, p. II-2 through II-7.
5. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards*- VOC Emissions from
Volatile Organic Liquid Storage Tanks —Background
Information for Proposed Standards. EPA-450/3-81-003a.
Research Triangle Park, NC. July 1984. p. 3-2.
6. Ref. 2, p. V-2.
7. Ref. 5, p. 8-2.
8. Ref. 2, p. II-8.
9. Memorandum from Probert, J.A., Radian Corporation, to RON
project file. February 3, 1992. Development of target
tank size distribution.
10. Memorandum from Probert, J.A., Radian Corporation, to
Kissell, M.T., EPA/SDB. September 14, 1992. Development
of Model Tank Farms for the HON Analysis.
11. Ref. 5, p. 4-11.
12. Schweitzer, P.A., ed. Corrosion Resistance Tables.
Second Edition. New York, Marcel Dekker, Inc. 1986.
pp. 6, 8, 418, and 442.
4-21
-------
-------
5.0 WASTEWATER
Facilities within the SOCMI have the potential to
generate wastewaters containing high concentrations of
organics, including HAP's. These wastewaters typically pass
through a series of collection units and primary and secondary
treatment units, which remove a portion of the organics. Many
of these collection and treatment units are open to the
atmosphere, which allows organic-containing wastewaters to
come in contact with ambient air. Atmospheric exposure of
these organic-containing wastewaters creates the opportunity
for volatilization of VOC's and HAP's.
Some emissions can be decreased through waste
minimization techniques, which reduce organic loading of the
wastewaters, or through waste stream segregation or recycling,
which reduces the quantity of wastewater generated. However,
some wastewater streams will still be generated. Emissions
from these streams can be reduced by installing add-on control
devices at the points of generation or by collecting
wastewater for treatment through an enclosed collection system
that is controlled for air emissions. Numerous controls are
suitable in specific cases, but the most universally
applicable technology for controlling emissions from
wastewater generated by facilities within the SOCMI is steam
stripping. This emission control method will-be discussed in
detail later in the text.
This section presents a discussion of the potential
sources of organic HAP and VOC emissions from SOCMI wastewater
streams. Section 5.1 describes the sources of organic-
containing wastewater and the sources of air emissions from
these wastewater streams. The model streams developed to
5-1
-------
represent the organic-containing wastewater streams generated
by the SOCMI are presented in Section 5.2. The environmental
and energy impacts of controlling these model emission sources
are presented in Section 5.3, and the cost impacts are
presented in Section 5.4.
5.1 EMISSION SOURCE DESCRIPTION
In the manufacture of organic chemicals, wastewater
streams containing organic compounds are generated. These
wastewaters are collected and treated in a variety of ways.
Some of these collection and treatment steps allow the
emission of organic HAP's and VOC's from the wastewater to the
air; this section provides a discussion of potential sources
of organic HAP and VOC emissions. Wastewater sources are
discussed in Section 5.1.1. Potential sources of HAP and VOC
emissions during wastewater collection and treatment are
discussed in Section 5.1.2.
5.1.1 Organic-Containing Wastewater
Many of the chemical processes in the SOCMI use organic
compounds as raw materials, solvents, catalysts, and
extractants. In addition, many of these processes generate
organic by-products during reaction steps. Consequently, many
of the wastewater streams that are generated by the targeted
product processes in the SOCMI are similar in organic content.
These organic-containing wastewater streams result from
(1) the direct contact of water with organic compounds during
chemical processing and (2) contamination of "indirect-
contact" wastewater through equipment leaks (i.e., wastewater
that is not intended to come in contact with organic compounds
in the process equipment but becomes contaminated with organic
compounds through equipment leaks).
Water comes in direct contact with organic compounds
through many different chemical processing steps and results
in wastewater streams that must be discharged for treatment or
disposal. Direct-contact wastewater includes:
• Water used to wash impurities from organic products
or reactants,
Water used to cool or quench organic vapor streams,
5-2
-------
Condensed steam from jet eductor systems pulling
vacuum on vessels containing organics,
Water from raw material and product storage tanks,
Water used as a carrier for catalysts and
neutralizing agents (e.g., caustic solutions), and
Water formed as a by-product during reactions.
Another source of direct-contact wastewater is water used
in equipment washes and spill cleanups. This wastewater is
normally more variable in flow rate and concentration than
those streams previously discussed, and it may be collected
for treatment differently from the wastewater streams
discharged from process equipment such as scrubbers,
decanters, evaporators, distillation columns, reactors, and
mixing vessels. •
Indirect-contact wastewater can become contaminated as a
result of leaks from heat exchangers, condensers, and pumps.
These noncontact wastewaters may be collected and treated
differently than direct-contact wastewaters. Pump seal water
is usually collected in area drains that tie into the process
wastewater collection system. This indirect-contact
wastewater is then combined with direct-contact wastewater and
transported to the wastewater treatment plant. Wastewater
contaminated- from heat exchanger leaks is often collected in
different systems and may bypass some of the treatment steps
used in the treatment plant. The organic content in these
streams can be minimized by implementing an aggressive leak
detection program.
5.1.2 Air Emissions
Wastewater usually passes through a series of collection
and treatment system units before being discharged from a
facility. Many of these units are open to the atmosphere and
allow organic-containing wastewaters to come in contact with
ambient air, thus creating an opportunity for organic HAP and
VOC emissions. The organic pollutants volatilize in an
attempt to reach an equilibrium with the vapor phase above the
wastewater. These organic compounds are emitted to the
ambient air surrounding the collection and treatment units.
5-3
-------
The magnitude of emissions depends on factors such as the
physical properties of the pollutants, the temperature of the
wastewater, and the design of the individual collection and
treatment units.
Collection and treatment schemes for wastewater are
facility-specific. The flow rate and organic composition of
wastewater streams at a particular facility are functions of
the processes used and influence the sizes and types of
collection and treatment units that must be employed.
Table 5-1 lists the potential sources of emissions in facility
collection and treatment systems at SOCMI facilities. The
following sections briefly discuss each of these emission
sources. A detailed discussion of each emission source,
including diagrams, typical design parameters, emission
mechanisms, and factors affecting emissions, is contained in
the Control Technology Center (CTC) document.1 In addition,
emission estimation models and example calculations for VOC
emissions for each source are presented in Appendices A and B
of the same document.
5.1.2.1 Drains. Wastewater streams from various sources
throughout a given process are introduced into the collection
system through process drains. Individual drains usually
connect directly to the main process sewer line, but may also
drain to trenches, sumps, or ditches. Some drains are
dedicated to a single piece of equipment, while others, known
as area drains, serve several sources. Many of these drains
are open to the atmosphere; that is, they are not equipped
with a water seal pot or p-trap to reduce the emission of
organic compounds to the atmosphere.
5.1.2.2 Manholes. Manholes are service entrances into
process sewer lines that permit inspection and cleaning of the
sewer line. They are placed at periodic lengths along the
sewer line or where sewers intersect or change significantly
in direction, grade, or line diameter. A typical manhole
opening is about 2 ft in diameter and covered with a heavy
cast-iron plate that contains two to four holes so that the
manhole cover can be more easily grasped for removal.
5-4
-------
TABLE 5-1.
EMISSION SOURCES IN WASTEWATER COLLECTION AND
TREATMENT SYSTEMS
Wastewater Collection System;
Drains
Manholes
Trenches
Sumps
Junction boxes
Lift stations
Wastewater Treatment Units;
Weirs
Oil/water separators
Equalization basins or neutralization basins
Treatment tanks
Biological treatment basins
Clarifiers
Surface impoundments
5-5
-------
5.1.2.3 Trenches. Trenches are used to transport
wastewater from the point of process equipment discharge to
wastewater collection units. In older plants, trenches are
often the primary mode of wastewater transportation in the
collection system. Trenches are often interconnected
throughout the process area and handle equipment pad water
runoff, water from equipment wash down and spill cleanups, and
process wastewater discharges. " Trench length is determined by
the locations of the process equipment and the downstream
collection system units, and typically ranges from 50 to
500 ft. Depth and width are dictated by the flow rate of the
wastewater discharged from process equipment and must be
sufficient to accommodate emergency wastewater flows from the
process equipment. Trenches are typically open or covered
with grates.
5.1.2.4 Sumps. Sumps are used to collect and equalize
wastewater flow from trenches before treatment. They are
usually quiescent and open to the atmosphere. Sumps are sized
based on the total flow rate of the incoming wastewater
stream.
5.1.2.5 Junction Boxes. A junction box combines
multiple wastewater streams into one stream which flows
downstream. Generally, the flow rate from the junction box is
controlled by the liquid level in the junction box. Junction
boxes are either square or rectangular and are sized based on
the total flow rate of the entering streams. Junction boxes
are typically open, but may be closed (for safety) and vented
to the atmosphere.
5.1.2.6 Lift Stations. A lift station is normally the
last collection unit before the treatment system, and accepts
wastewater from one or several sewer lines. The main function
of the lift station is to collect wastewater for transport to
the treatment system. A pump provides the necessary head
pressure for transport and is usually designed to switch on
and off based on preset high and low liquid levels. Lift
stations are typically rectangular in shape and greater in
5-6
-------
depth than length or width and are either open or closed and
vented to the atmosphere.
5.1.2.7 Weirs. Weirs act as dams in open channels. The
weir face is usually aligned perpendicular to the bed and
walls of the channel. Water from the channel normally
overflows the weir but may pass through a notch, or opening,
in the weir face. Because of this configuration, weirs
provide some control over the level and flow rate through the
channel. Weirs may also be used for wastewater flow rate
measurement. Water overflowing the weir may proceed down
stair steps, which aerates the wastewater. This increases
diffusion of oxygen into the water, which may benefit the'
biodegradation process (often the next treatment step).
However, this increased contact with air also accelerates the
volatilization of organic compounds contained in the
wastewater.
5.1.2.8 Oil/Water Separators. Oil/water separation is
often the first step in wastewater treatment, but oil/water
separators may also be found in the process area. These units
separate and remove oils, scum, and solids from the wastewater
by gravity. Most of the separation occurs as the wastewater
stream passes through a quiescent zone in the unit. Oils and
scum with specific gravities less than water float to the top
of the aqueous phase, while heavier solids sink to the bottom.
Some of the organic compounds contained in the wastewater will
partition to the oil phase and then can be removed with the
skimmed oil, leaving the separator.
5.1.2.9 Equalization Basins. Equalization basins are
used to reduce fluctuations in the temperature, flow rate, and
organic_ compound concentrations of the wastewater going to the
downstream treatment processes. The equalization of the
wastewater flow rate results in more uniform effluent quality
from downstream units and can also benefit biological
treatment performance by damping any influent concentration
and flow fate fluctuations. This damping protects biological
processes from upset or failure caused by shock loadings of
toxic or treatment-inhibiting compounds. Equalization basins
5-7
-------
normally use hydraulic retention time to ensure equalization
of the wastewater effluent leaving the basin. However, some
basins are equipped with mixers or surface aerators to enhance
the equalization, accelerate wastewater cooling, or saturate
the wastewater with oxygen before secondary treatment.
5.1.2.10 Treatment Tanks. Several different types of
treatment tanks may be used in wastewater treatment systems.
Tanks designed for pH adjustment are typically used preceding
the biological treatment step. In these tanks, the wastewater
pH is adjusted, using acidic or alkaline additives, to prevent
shocking the biological system downstream. Flocculation
tanks, on the other had, are usually used to treat wastewater
after biological treatment. Flocculating agents are added to
the wastewater to promote formation or agglomeration of larger
particle masses from the fine solids formed during biological
treatment. These larger particles precipitate more readily
out of the wastewater in the clarifier, which usually follows
in the treatment system.
5.1.2.11 Biological Treatment Basins. Biological waste
treatment is normally accomplished using aeration basins.
Microorganisms require oxygen to carry out the biodegradation
of organic compounds, which results in energy and biomass
production. "The aerobic environment in the basin is normally
achieved with diffused or mechanical aeration. This aeration
also maintains the biomass in a well-mixed regime. The
performance of aeration basins is particularly affected by
(1) mass of organics per unit area of wastewater,
(2) temperature and wind patterns, (3) hydraulic retention
time, (4) dispersion and mixing characteristics,
(5) characteristics of the solids in the influent, and
(6) amount of essential microbial nutrients present.
5.1.2.12 Clarifiers. The primary purpose of a clarifier
is to separate solids from wastewater through gravitational
settling. Most clarifiers are equipped with surface skimmers
to clear the water of floating oil deposits, grease, and scum.
Clarifiers also have sludge-raking arms that remove the
accumulation of organic solids that collects at the bottom of
5-8
-------
the tank. The depth and cross-sectional area of a clarifier
are functions of the settling rate of the suspended solids and
the thickening characteristics of the sludge. Clarifiers are
designed to provide sufficient retention time for the settling
and thickening of these solids.
5.1.2.13 Surface Impoundments. Surface impoundments are
used for evaporation, polishing, storage before further
treatment or disposal, equalization, leachate collection, and
as emergency surge basins. They may be quiescent or
mechanically agitated.
5.2 MODEL EMISSION SOURCE DEVELOPMENT
In developing a NESHAP regulating emissions of HAP's from
wastewater in SOCMI facilities, baseline nationwide emissions
and national impacts of various control options had to be
estimated. Where available, actual wastewater stream
information was used to estimate emissions, emission
reductions, and control costs. However, the detailed
information necessary for such an assessment was not available
for every facility. Therefore, for those process units where
actual wastewater stream data was incomplete, available data
was supplemented with engineering judgement. Representative
flows and concentrations were assigned to these wastewater
streams so baseline emissions and impacts could be estimated.
Impacts based on model wastewater streams that were developed
to represent the reported streams are presented in
Sections 5.3 and 5.4. The procedure used to develop these
model wastewater streams is described in the following
sections.
5.2.1 Data Gathering
Under the authority of Section 114 of the CAA, nine
corporations were asked to provide information on their SOCMI
chemical processes. The list of corporations to be surveyed
was developed to maximize the number of chemical processes for
which information could be obtained. General information was
requested for all the SOCMI chemical processes (as defined by
40 CFR Part 60, Subpart W)2 that use or produce HAP's at each
facility. These same corporations were asked to provide
5-9
-------
additional information on wastewater generated by a subset of
the reported processes. Such information was requested for
120 process units at a total of 27 of the original
29 facilities surveyed.3
The information on wastewater processes included a
general process description; a process block flow diagram
identifying processing steps, product streams, arid wastewater
streams; and information on all wastes and intermediate
materials that contain candidate HAP's. For each wastewater
stream with an average flow rate greater than 0.1 gpm (or
total annual flow greater than 5,000 gpy) , the fcicility was
asked to provide information on the flow rate and the
concentration of individual HAP's and total VOC's. Other
information requested included design and performance data on
any organic compound recovery/removal operations conducted on
wastewater streams at the facility, and influent and effluent
data for the combined facility wastewater treatment system.
The individual waste stream information gathered from the
Section 114 questionnaires was entered into a data base for
analysis. This data base contains information ort
25 facilities and 110 process units that reportedly produce a
total of 461 wastewater streams containing HAP's.3
Other information used to develop model wastewater
streams included data gathered in facility visits;. Emission
factor estimation procedures, control cost estimates, and
control effectiveness estimates presented in the CTC document1
were also used to estimate baseline emissions and control
impacts.
5.2.2 Model Development-
To represent the range of impacts associated with these
wastewater streams, model wastewater streams were developed
from the Section 114 data base. This section presents the
rationale for the model stream parameters selected and also
presents the values chosen for the selected parameters to
represent the wastewater streams generated within the SOCMI
processes.
5-10
-------
Three parameters were identified as having the greatest
impact on emissions, emission reductions, and costs:
wastewater flow rate, HAP concentration, and HAP volatility.
These parameters are discussed in Table 5-2, along with the
methodology used for model calculations.
While creating the Section 114 data base, values for
these three parameters were examined from the reported data
for each wastewater stream. Based on these individual stream
data, model wastewater streams were created to span the ranges
of flow rates, HAP concentrations, and HAP volatilities
present in the data base. To represent these ranges, seven
flow rate ranges, three HAP concentration ranges, and four
volatility ranges were selected, based on an engineering
4
review of trends in the data. From each range, a value was
selected to represent that range. As shown in Table 5-3, this
process created a total of 84 combinations of flow rate, HAP
concentration, and volatility.
5.2.3 Model Characteristics
Emissions of HAP's from a wastewater stream are a
function of the wastewater stream flow rate, the HAP
concentration in the wastewater stream, and the predicted
fraction emitted (which is a function of volatility). The
emission reduction that is achievable through treatment of the
wastewater streams by the design steam stripping system
presented in Section 2.2.3 of BID Volume IB is a function of
the HAP concentration, the wastewater stream flow rate, as
well as the predicted fraction emitted and the predicted
strippability (which are functions of volatility) for the
organic HAP's present in the wastewater stream. The cost of
controlling wastewater streams by steam stripping and
controlling the organics removed from the wastewater is
primarily a function of the wastewater stream flow rate.
Although predicted impacts were calculated based on the
specific stream characteristics, a subset of 18 examples were
selected from the 84 model wastewater streams to illustrate
the possible HAP and VOC emission reductions and cost impacts
that could result from treating wastewater streams using steam
5-11
-------
in
tr
o
Q)
O
rH
a
u
a)
a
ft
•H
W
ft
(0
Q)
tn
G
O J-l
•H 0)
-p -p
(d (d
-P <
a)
4->
CQ
O
(d
ft
•H-H fi
g -O-H
0)
•H
to
-P -P
(0 CO
10
J ^
<4-l Ft. CM
O rt!
0)
tQ -P tO g
CM
to en
(0 Q) O
0) g
(0
-P 0)
(d (0
O IS 4J
10
tn 43
(d -P a)
O 0)
.p 4J c
0) O
a) -P
4-> O
to c
><(«
a)
-H
0)
tn fe
•H H a,
4J <
d "aT>—
•d c
0) -H |XJ
^ rH
Q) \
c tn —
O rtl -H
•H 43 M
CO b
to c
•HO*
e-H
0) W -H
Ul 0)
&>-H Pn
g g
•^ QJ *
Oi
Q)
Ou
ft 0)
•H C
^4 -H
-P rH
W
•P
•H tn o
Q) > -H C
43-H g 3
EH T3 Q) t-l
-H
5-)
O
to
43
3
to
0)
0)
Q)
Q)
^
(0
cn
O
(0
0)
0)
to
(D
(0
>
(0
5-12
-------
TABLE 5-3. HON WASTEWATER MODEL STREAMS
Model
Stream
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
. Flow
[(£pm)/(Gg/yr)]
0.005
0.005
0.005
0.05
0.05
0.05
0.15
0.15
0.15
0.5
0.5
0.5
1.5
1.5
1.5
10
10
- 10
250
250
250
0.005
0.005
0.005
0.05
0.05
0.05
0.15
0.15 '
0.15
0.5
HAP Concentration
(mg/£)
10
250
8000
10
250
5000
10
250
5000
10
250
3000
10
200
1600
10
200
1600
10
200
1600
10
250
8000
10
250
5000
10
250
5000
10
Volatility3
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
5-13
-------
TABLE 5-3.
HON WASTEWATER MODEL STREAMS
(CONTINUED)
Model
Stream
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
Flow
[(£pm)/(Gg/yr)]
0.5
0.5
1.5
1.5
1.5
10
10
10
250
250
250
0.005
0.005
0.005
0.05
0.05
0.05
0.15
0.15
0.15
0.5
0.5
0.5
1.5
1.5
1.5
10
10
10
250
HAP Concentration
(mg/£)
250
3000
10
200
1600
10
200
1600
10
200
1600
10
250
8000
10
250
5000
10
250
5000
10
250
3000
10
200
1600
10
200
1600
10
Volatility3
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
Medium-High
5-14
-------
TABLE 5-3.
HON WASTEWATER MODEL STREAMS
(CONCLUDED)
Model
Stream
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
Flow
[(£pm)/(Gg/yr)]
250
250
0.005
0.005
0.005
0.05
0.05
0.05
0.15
0.15
0.15
0.5
0.5
0.5
1.5
1.5
1.5
10
10
10
250
250
250
HAP Concentration
(mg/4)
200
1600
10
250
8000
10
250
5000
10
250
5000
10
250
3000
10
200
1600
10
200
1600
10
200
1600
Volatility51
Medium-High
Medium-High
High
High
High
High
High
High
High
High
High
High
_> High
High
High
High
High
High
High
High'
High
High
High
aFor volatility
For volatility
For volatility
For volatility
Low, Fe = 0.03 and Fr = 0.05
Medium, Fe = 0.19 and Fr = 0.40
Medium-High, Fe = 0.25 and Fr = 0.70
High, Fe = 0.68 and Fr = 0.98
5-15
-------
stripping. These 18 example model wastewater streams (shown
in Table 5-4) were selected to provide a manageable number of
examples while still illustrating the full range of impacts.
Wastewater stream flow rates are presented in this table in
£pm/Gg/yr production capacity and in £pm (assuming a process
unit capacity). Because impacts are a function of the
wastewater stream flow rate (£pm), which is the product of the
model stream flow rate (£pm/Gg/yr) and the capacity (Gg/yr) of
the process unit, example process unit capacities were assumed
for this illustration of impacts." The capacities that were
selected produced wastewater stream flow rates (£pm) typical
of those in the Section 114 data base.
When calculating treatment costs, it was assumed that
facilities would combine wastewater streams for treatment
whenever technically feasible. Based on this assumption, the
total flow treated by the steam stripper was included as an
additional parameter for evaluating cost impacts. Steam
stripper feed rates of 50 and 500 £pm (10 and 130 gpm) were
selected as examples for choosing the appropriate-sized steam
stripper.6 Unit treatment costs ($/yr/£pm) were then
calculated for the steam stripper and applied to the
individual streams.
High, medium-high, and low values for volatility and HAP
concentration and high and low flow rates were selected to
represent the range of uncontrolled emissions, emission
reductions, steam stripper costs, and cost effectiveness. The
HAP and VOC emission reductions achieved through steam
stripping by the design steam stripping system presented in
Section 2.2.3 of BID Volume IB depend on the HAP
concentration, the volatility, and the wastewater stream flow
rate.
5.3 ENVIRONMENTAL AND ENERGY IMPACTS OF CONTROLLING EMISSIONS
FROM WASTEWATER
The purpose of this section is to evaluate the
environmental and energy impacts associated with steam
stripping. Steam stripping is a control technique that
removes organic compounds from wastewater before the
• 5-16
-------
TABLE 5-4. EXAMPLE MODEL WASTEWATER STREAMS
Model
Stream
4
5
6
46
47
48
67
68
69
19
20
21
61
62
63
82
83
84
Stream Flow
(£pm/Gg/yr)
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
250
250
250
250
250
250
250
250
250
Stream
Flowa
(£pm)
10
10
10
10
10
10
10
10
10
100
100
100
100
100
100
100
100
100
Facility
Flowb
(£pm)
50
50
50
50
50
50
50
50
50
500
500
500
500
500
500
500
500
500
HAP
Cone.
(mg/£)
10
250
5000
10
250
5000
10
250
5000
10
200
1600
10
200
1600
10
200
1600
Volatility0
Low
Low
Low
Medium-High
Medium-High
Med ium-H igh
High
High
High
Low
Low
Low
Medium-High
Medium-High
Medium-High
High
High
High
aBased on responses to Section 114 questionnaires. Assumed
capacities of 200 Gg/yr for the 0.05 £pm/Gg/yr models, and
0.4 Gg/yr for the 250 £pm/Gg/yr models.
bBased on responses to Section 114 questionnaires. Assumed
facility treated flow rates to be 50 £pm for the 10-£pm
models, and 500 £pm for the 100-£pm models.
GFor Volatility = Low, Fe = 0.03 and Fr = 0.05,
For Volatility = Medium-High, Fe =0.25 and Fr = 0.70
For Volatility = High, Fe = 0.68 and Fr = 0.98
5-17
-------
wastewater contacts ambient air. The recovered organic
compounds may be returned to the process or may be burned as
fuel in a combustion device. Steam stripping effectively
reduces HAP and VOC air emissions that occur during downstream
wastewater collection and treatment and also improves water
quality. Analysis of the environmental impacts of this
control technique included an evaluation of air and water
pollution, waste disposal, pollution prevention, and energy
use. Section 5.3.1 presents an assessment of primary air
pollution impacts, including HAP and VOC emission impacts;
Section 5.3.2 covers secondary air pollution impacts resulting
from fuel combustion for production of steam; and 5.3.3
discusses water pollution, solid waste, pollution prevention,
and energy impacts.
5.3.1 Primary Air Pollution Impacts
The reduction in HAP and VOC emissions that can be
achieved by steam stripping a wastewater stream is dependent
on the stripper design and the characteristics of the
wastewater streams (i.e., flow rate, composition, and
concentration). Table 5-5 presents HAP and VOC emission
reductions achievable through steam stripping for the example
model wastewater streams listed in Table 5-4. Also presented
in Table 5-5' are baseline emissions and controlled emissions
for each model wastewater stream.
5.3.2 Secondary Air Pollution Impacts
This section evaluates the on-site secondary emissions
associated with steam stripping. These secondary emissions
are compared to HAP and VOC emission reductions for the
18 example model wastewater streams.
Secondary air impacts can occur from two sources:
(1) combustion of fossil fuels for steam and electricity
generation, and (2) handling or combustion of the recovered
organic compounds. For the purpose of this evaluation, it is
assumed that recovered organics are handled properly and
either returned to the process or combusted. Fuel combustion
for steam and electricity generation is a source of combustion
pollutants—particulate matter (PM), sulfur dioxide (SO2),
5-18
-------
co
SH
rn
s
Cj
CO
w
EH
g
W
E-<
CO
g.
w
Q
0
g
p.
3
ft
§
S
w
o
CO
o
H
B
B
§
S5
o
H
CO
CO
H
g
W
u
Q
S
04
S
^
in
in
E3
ffl
g
o
w
^
o
•H
4->
O
3
0]
tt
sion
•H
w
^
rH
O
i 1
4-*
O
u
OJ
<
en
c
-H
Ul
CQ
•H
g
w
(0
CO
c
o
•H
U
CO
•H
s
w
(U
c
•H
, — |
a)
en
aa
^
u
^>
E
TT
>^
o7
hH
^
g
**-*
^^
J^
^1
O
Cr
—
^^
^
\
§
E
^^
^
>«
>
£
^"*
j_i
\
Oj
ffi
g
rH |
I
s w
CO CTi
co r^ *^* CM co
OOHOWCT\rHCM'*OO<3 r- — H
rH r^ co c^ o co in CTV r*"
,OCOinOCT> Ot-HVOHO COO^ O'l'
OOOOOCMOOOOOCMOOVOOOH
CO
in co co
o in H <5j* co in co
coroTr^^'coco^4
(__„_, ^,_^_,_.
CM AO r-~ CM H
CM
vo
H G\ CO VO VO
OfOO^rHfO COCTi rHCM CO VO
OOt^OCOVOOCOCTiOCOinHVOOCOrHfM
OOOOOVOOOC~-OOCMOCMHO[^-r-
I-H CM in
"««5i;3{SSS.SSS5SSgSS
5-19
-------
in
i
in
*
tn
g"
cr>
o
H
*
"i"
^
.
15 £-1
O ><
Fr. tj
ir*
*
o
^-^ vo
^ r**
^x. CO
CT>
& *
v_*
£_J
pj 1-*
i^3 ^x
a cs
•H
* g
0) 0
fa vo
II
0>
CO
C
o
•H
CO
CO
•H
g
W
CM
a
0)
C
•H
H
Q)
CO
id
CQ
II
&>
CO
C
missio
w
u
o
(D
C
•H
rH
•Q)
CO
(d
CQ
where :
•.
n3
•H C!
rH O
•H -H
Q ^J
td cd
C^i ^
a-p
•H C
S-l 0)
•p o
CO C
— ' O
o
*$
C
O (d
0) fi -P
-P Q) S
cd JH i — i
^ rH
, CQ O
^ co a
O (d
rH £ J-l
(w (d
M O
(D CO
-PCS
(d o o
[5 -H T3
0) 4-1 ^1
4-> O (d
CQ (d N
(d M id
15 1
1
C
o
•H
4->
O
n
Ti
Q)
DH
CO
o
•H
CO
CO
• -H
&
U
1
^^
j_l
>.
\
E
CO
q;
O
-H
CO
CQ
-H
g
H
CD
g
-rH
rH
0)
CO tO
vo
to
^
"$7
>i
d?
s
CO
0
nissi
w
a)
c
•H
rH
0)
CQ
CO
II II II II
o
•H
(d
.., CQ CO
1 a,§§
0) rH >-l < .^ -H
CO CO
•H -H
S g
W W
U CM
O a
II CQ
— u
CQ
C
O
•H
CQ
CQ
•H
g
Q)
4J
C
O
U
X!
o
•H
-P
O
Q)
C
O
•H
CQ
CQ
o
5-20
-------
NOX, CO, and VOC's. Only steam generation was assumed to
occur on site; therefore, impacts from off-site electricity
generation are not discussed here.
The secondary emissions presented in this section were
estimated using EPA emission factors presented in Table 5-6.
Assumptions concerning the fuel composition and boiler
efficiency are based on information compiled by the EPA and
7 8
the Energy Information Administration. ' These values were
adjusted to accommodate emission reductions by existing
control devices. Typical controls and control efficiencies
presented in these sources were assumed.
The industrial boiler used for steam generation was
assumed to have a capacity of less than 158 MMkJ/hr
(150 MMBtu/hr). An efficiency of 80 percent was assigned to
the industrial boiler as an average expected value. It is
assumed to be controlled for SO2, PM, and NOX emissions using
desulfurization (90 percent SO2 removal efficiency), an
electrostatic precipitator (99 percent PM removal efficiency),
and flue gas recirculation (assuming the mid-range of
9 10
40 percent NOX removal efficiency), respectively. ' Fuel
composition was based on national fuel use for industrial
boilers: natural gas at 45 percent, residual oil at
28 percent, distillate oil at 7 percent, and coal at
20 percent. Average heating values are presented in
Table 5-6.
Estimated emissions, based on these assumptions were
calculated as follows:
Uncontrolled Emissions = Annual Fuel Use * Emission Factor
and
Controlled Emissions = Uncontrolled Emissions * (1 - Control
Efficiency)
The resulting secondary emission estimates for the 18 example
model streams are presented in Table 5-7. Figure 5-1 presents
normalized secondary air impacts (Mg/yr) from a controlled
fossil fuel boiler generating steam for steam stripping
wastewater streams. The steam requirements were based on the
steam stripper design presented in Section 2.2.3 of BID
5-21
-------
(d
o
H
EH
r5
W
H
H
o
s
W
CO
rv*
M4
O
CO
55
f"_i
EH
O
"3
Cti
,£4
O
H
CO
CO
H
H
EH
§
EH
i-3
0
O^4
H
EH
CO
§
S
O
o
•
vo
1
in
E
«
EH
0)
j3
rH
fS
CT
•H
4->
fd
0)
O
0
0
r_>
^
O
a
o
O
co
s
co CM t^ in
o o o o
H H H H
* * * *
co o CM in
• • • •
CO H •* H
CO
CM
«3* O
co o
CM *# O O
CM rH O 0
O
in
O O
VO O
^ o o o
vo *3*
in
in
VO O
o in vo o
•* CM
CM VO
O VO OH
• • • •
HO CTI o
H
CO
H
CO O VO O
«3* CO HO
^"*
co 4-J
g CQ
a a)
!3 Pn
r- in
o o
H H
* *
CO ^
• •
CO H
O
CM
••* O
CM O
0 0
O O
O
in
o o
VO O
, .
o o
o
CM
^ O
CM O
"3*
0 H
r- o
H
CM
<* o
CM O
O 0
H
-H
o —.
Q) x-^H *-~. (0
-Pco (dco tri
a & &&-\
H \\\ 3
H C^X! l-q ±)
•H x H £ m
-H
Q
•* «*
o o
H H
* *
CO CM
• •
CM H
in
co
o
in o
CO O
o o
• *
0 0
o
CO
o
o o
CO O
, 9
o o
H
H
O O
H O
j—j
(71
Tl CM
O O
cy> o
CM
O vo
O 0
o o
vo
i — I
(d
o
u
-d -^ — xT
0) --^Xt (T>H
N tn -P
0) CTH ^ ffl
rH
o
X!
!5 rH
(d
o
rH O
(d
o c;
O -H
C SH
•H 3
4J > 0)
•H c -H a
CnxD Cn
o m
•H O 0) O H
O 4-> &4->
0 O 4->
H (d CM (d (d
3 Q)
CO 3 CO 1
•H rH -H TJ
C g (d g -H
•H CU > Q) g
4-> -H
c cu a Q) tX! X!
O 4-1 4-> 4-)
O
M g (d g 4->
• 3 O O
H > CO
CU O -H T3 -H (d
(U H 0) g CU CO
CU CO CO S
^ a) !>-i co M
g 0 (0 0 X3
g 3 4-> 4-) O
O CO O CO O -H
J-) co (d (d (d x!
&H i< fa £ PL, |3
(d X! O T5
5-22
-------
TABLE 5-7. SECONDARY AIR POLLUTION IMPACTS OF EXAMPLE MODEL
WASTEWATER STREAMS3
Controlled Pollutant Emissions
(Mg/yr)
Model
Stream
PMb S02b
CO
voc
4, 5, 6,
46, 47, 48,
67, 68, 69
19, 20, 21,
61, 62, 63,
82, 83, 84
0.006 0.05
0.06 0.5
0.15
1.5
0.02
0.2
0.001
0.01
aFuel composition for steam generation is based on 45, 28, 7,
and 20 percent natural gas, residual oil, distillate oil, and
coal, respectively.
, NOX, and PM controls reduce emissions by 90, 40, and
99 percent, respectively.
5-23
-------
°8
o>
CVIQ
do
rH
o
o
o
0)
ft
-P
U
0)
-P
g
O
0)
-------
Volume IB. The secondary air impacts were normalized for a
treated wastewater flow rate of 100 £pm. Figure 5-2 compares
the normalized secondary NOX emissions to a range of
normalized VOC and HAP emission reductions achievable from
steam stripping wastewater streams with variations in
concentration and strippability of organic compounds, as
represented by model streams 19, 62, and 84.
Handling the recovered organics for disposal may also
contribute to secondary air impacts. For example,
incineration of recovered organic compounds produces
combustion pollutants as a secondary impact. However, the
recovered organic compounds could be used as an alternate
energy source, that is, to generate some of the steam required
by the steam stripper. Although combustion of the organic
compounds will produce combustion pollutants, the emissions of
S02 and PM will typically be less than those generated by
fossil fuel combustion. This is due primarily to two factors:
(1) most organic compounds do not contain sulfur, which reacts
to form SO2 when burned, and (2) organic compounds do not
contain high concentrations of inorganics, which are emitted
as particulates when burned. If recovered organic compounds
are recycled (i.e., not combusted), then they do not
contribute to the secondary air impacts.
5.3.3 Other Impacts
5.3.3.1 Water Pollution Impacts. Because steam
strippers remove organic compounds from the wastewater,
thereby improving the quality of wastewater being discharged
to the wastewater treatment plant or to a POTW, their use has
a positive impact on water pollution.
5.3.3.2 Solid and Hazardous Waste Impacts. Solid and
hazardous waste can be generated from three possible sources:
organic compounds recovered in the steam stripper overheads
condenser, solids removed during feed pretreatment, and wastes
generated in the control of system vent emissions. System
vent emissions, if not sent to a combustion control device,
may be collected on a sorbent medium that requires either
5-25
-------
»a
crc
X
X
X
IX
1
A C
Pi* , *!
i i II t *
i „ - >, c |
CM
CO
£JO>T-
CD T— CO
pi JJQ-O
'ft' ~~ $ ^
jl:" •§
=J 2
o>
T—
ro o O
^ 1^1 C^
1^ |
3
T3
1
CO
111 L
o c
5 5
HI
• i ' i ' i • i * i ' i ' i '
cS
ECON.
2c\ico
C^Q-O
|S§
5
•5 CD
D |
^ 1
| «
1 |
CO
E 1
^
rH
O
-P
C
o
o
0)
a.
-P
u
g
-------
disposal or regeneration. If the sorbent is disposed of, it
creates additional solid waste.
Although waste generation can increase for any
nonrecyclable organics that cannot be used as supplemental
fuel, these organic wastes most likely would have been removed
otherwise from the wastewater via the air (volatile organics
only) or via an oil/water separator. Similarly, solids
removed from the wastewater in cases where pretreatment is
necessary would have likely been removed in a clarifier or
activated sludge unit.
5.3.3.3 Pollution Prevention. As described in
Section 2.2.3 of Volume IB, the condenser unit in a steam
stripping system is used to recover the organic and water
vapors in the overheads stream. The organics recovered are
usually either pumped to storage and then recycled to the
process or burned as fuel in a combustion device such as the
steam-generating boiler.
For organic compounds recovered for use in the
manufacturing process, there is a reduction in the amount of
raw materials that must be used in the process. Another
option for recovered organic compounds is to sell them to a
chemical manufacturer who will recover the separate components
of the waste organic compound stream.
If the recovered organics are burned as fuel in a
combustion device, there is a net benefit to the environment
due to reduction in the usage of conventional fuels. However,
this approach would not be considered a pollution prevention
measure.
5.3.3.4 Energy Impacts. The additional fuel demand to
generate steam for the steam stripper system reduces available
nonrenewable resources: coal, oil, and natural gas. This can
be partially offset if the recovered organics are used as
supplementary fuel or if they are recycled. (Recycling
reduces the facility demand for petroleum-derived feedstocks.)
Table 5-8 summarizes the annual fuel usages for steam
generation for two example model streams. These model streams
were selected to indicate the range of annual fuel usage that
5-27
-------
TABLE 5-8. ANNUAL FUEL USE FOR STEAM GENERATION FOR
STEAM STRIPPER CONTROL OF EXAMPLE MODEL STREAMS3
Model
Stream
Fuel
Percent
Compos it ion*3
Annual Use
Natural gas
Residual oil
Distillate oil
Coal
45
28
20
1.59 * 104 m3
(5.63 * 105 ft3)
9.27 m3
(2.45 * 103 gal)
2.50 m3
(6., 57 * 102 gal)
9.48 * 103 kg
(2.09 * 104 Ib)
18 Natural gas
Residual oil
Distillate oil
Coal
45
28
20
1.59 * 105 m3
(5.63 * 106 ft3)
9.27
m-
(2., 45 * 104 gal)
2.50 * 101 m3
(6 ,,57 * 103 gal)
9.48 * 104 kg
(2.09 * 105 Ib)
aBased on steam stripper design in Section 2.2.3 of BID
Volume IB.
bBased on national fuel use for industrial and electrical
generating boilers.
5-28
-------
can be expected if air emissions from wastewater are
controlled with a steam stripper. These values are based on
the steam stripper design presented in Section 2.2.3 of BID
Volume IB and the boiler capacity and efficiencies discussed
previously. The fuel composition assumed for steam generation
is as follows: 45 percent natural gas, 28 percent residual
oil, 7 percent distillate oil, and 20 percent coal. These
•
percentages were based on national fuel use data for
industrial boilers.
5.4 COST IMPACTS OF CONTROLLING EMISSIONS FROM WASTEWATER
Facilities using steam stripping to remove organic
compounds from wastewater and thus reduce the potential for
air emissions of HAP's and VOC's will likely not apply a
separate steam stripper to each individual wastewater stream.
Facilities will more likely combine wastewater streams
whenever possible for more economical treatment. Therefore,
the cost impacts of steam stripping are dependent on these
combined stream flow rates. The range of capital and annual
cost impacts to a facility using steam stripping is
illustrated using two model combined stream wastewater flows—
50 and 500 £pm (see Table 5-4). These two combined streams
illustrate the range of costs that could be incurred by a
facility when steam stripping its wastewater.
5.4.1 Capital Costs
Section 3.2.3 of Volume IB of this document describes the
methodology for calculating capital costs for steam stripping
systems. Appendix D of the same volume contains an example of
this methodology applied to a stripper design to treat a
500-£pm wastewater stream.
The base equipment cost for a steam stripping system
treating a facility flow of 50 and 500 £pm are given in
Tables 5-9 and 5-10, respectively.12'13'14'15'16'17 The total
base equipment cost then becomes the basis for the estimation
5-29
-------
55
H
ft
ft
H
g
W g
§ ^
Is
CO '
4J
in rcj
§S
u
Equipment
Size
Equipment
Component
f\j
rH
O
O
O
CO
0)
0)
Ul
C
O
•8
fO
o
CO
O
vo
H
cn
c
(0
TJ
0)
0)
OJ
0
o
o
cn
rH
0)
0)
4-J
in
0
(0
o
Cg
o
^_^
preheater
1 and tube
rH
T3 Q)
0) X!
Q) in
CO
rH
O
O
O
VO
rH
ffl U
» **^
•g g
rH
e G
O "H
LJ ~P
in
u
Q)
4J 4J
Q) X!
g CT1
(0 -rH
•H 0)
73 X!
H <7i
CO CM
O
stripping
column
g
(0 £*i
OJ n3
• \ i j
CO 4-1
in
H
o
o
o
co
r — 1
0)
Q)
4->
in
C
0
(0
o
CM
g
tj
0)
ry condens
r cooled,
and tube)
(0 Q) rH
g 4-> rH
•H rtl Q)
SH IS x!
ft ^- in
vo
cH
O
0
0
H
1
r~i
0)
Q)
4->
in
c
o
•a
a
CJ
CO
rH
O
C
O
•H
O
Q)
rH
i-H
O
O
IQ
(0 0)
Q) 4J
XI C
S-J (^j
0) O
> Q)
O 73
r- in
rH cH
• ! 1
H H
H
Q)
CD
4->
rn
UJ
Xi ifl
rfj in
>• fii
^— t *»*
r—j
C
•H
fd
^j
CO
«, Q) 4->
0 C S
1s "
S-y *
r c 4->
« 0) 0
<« > 4J
C ^ °
So> o
0 ft o
H
arrestor
(4)
o) cn
g PI
n3 g
rH 3
pTj p [
O
o
o
CO
o
H
EH
w
o
u
H
t)
a
w
w
o
EH
M
M
(0
co
CTi
(0
0)
rH
X!
(0
O
&
4J
O
5-30
-------
EH
H
O
S
H
ft'
H
«
EH
a ft
S ^
m S
^ fa
0 IS
u w
r_l
W W
n, BH
rl W
cx12
ATION OF BASIC
^TING A FACILITY
^ ls^
H W
Ej EH'
W
o
,__)
I
in
t4
m
IS
(U
o
-*-1 §
in R
Sfl
0)
•P
l\~
&s"
w
C
o
•H rH
-P fd
Construe
Materi
Equipment
Size
-P -P
c c
0) 0)
K C
ft O
•H ft
3 g
CT* o
w u
CM
H
-
o
O
0
cn
rH
Q)
0)
-P
Carbon s
Is
o
o
vo
H
tanks
73
0)
0)
fa
CM
H
O
O
o
o
in
rH
Q)
Q)
-P
Carbon s
CM
g
o
o
^^
0)
0) 3
-P 4J
Q) tJ
X! C
a) fd
ftH
rH
•d a)
0) x!
o in
fa ^"*
-*
r~'
n
H
o
o
o
in
xj
•rH
in
rH >1
rH fd
0) rH
X! -P
m
n steel £
tainless
o w
XI
i (d
rH rH
(d 0) rH
6 H-> rH
•H (0 0)
rH S: XI
ft — W
VO
H
O
0
O
rH
H
rH
0)
0)
-p
Carbon s
en
g
CO
o
rH
C
in o
T3 -H JH
(d -P (U
0) O -P
x! arc
rH rH (d
0) H O
> O 0)
o o -d
r-
rH
O
O
0
H"
X!
•z
One arrestor
per vent line
^
o
-P
to
arre
0)
fi
fO
rH
fa
in
H
o
0
o
vo
in
O
0
o
CO
CM
j
rH
W
in
M
0)
rH
C
•H
(d
-p
w
rH
(d
-P
0
•P -P
-P
o td
0 S
CO
o
rH
S
in
ft
rj
ft
r_,
o.
U
BASE EQUIPMENT
hi
EH
O
EH
.
0)
rH
m X!
rH (d
id o
rH -H
rH rH
0 ft
•C ft
fd
CTi
co -P
cr\ o
t^ 11
rH
3 <
l~3 S
(d XJ
5-31
-------
of total capital investment,18 shown in Tables 5-11 and 5-12,
for the model facility flows of 50 and 500 €pm. All costs
shown are in July 1989 dollars.
5.4.2 Annual Costs
The methodology for calculating annual costs for steam
stripping systems was also presented in Section-3.2.3 of
Volume IB of this document. Appendix D in the same volume
showed the application of this methodology to the example
500-£pm stripper. Tables 5-13 and 5-14 show the estimated
total annual cost for treating the example facility wastewater
flow rates of 50 and 500 £pm with a steam stripping system.
To calculate the capital recovery factor, a. 15-year lifetime
was assumed for all equipment and a 10-percent interest rate
was used.
The estimated annual unit operating costs for the steam
stripper system at 50 and 500 £pm are $0.0046/4 and $0.0016/4
of wastewater treated, respectively. The treated wastewater
costs have been estimated in July 1989 dollars.
To assess the impact of plant size, annual unit operating
costs were estimated for four other facility wastewater flow
rates: 40, 150, 455, and 760 4pm. Figure 5-3 illustrates the
results of these cost estimates for both carbon steel and
stainless steel construction. As can be seen from Figure 5-3,
a steam stripper of stainless steel construction is more
costly than one constructed of carbon steel. The figure also
shows that steam stripper system unit operating costs decrease
with increasing flow rate. However, unit operating costs are
fairly constant for wastewater feed rates greater than or
equaL to 300 4pm.
5.4.3 Cost Effectiveness
Cost effectiveness for the control of HAP emissions is
defined as the total annual control cost per megagram of HAP
emissions reduced. The cost effectiveness for the selected
18 model streams is presented in Table 5-15.
Estimates of cost effectiveness for both carbon steel and
stainless steel construction are illustrated in Figure 5-4 for
an example stream composition of 2,500 ppm VOC's at different
5-32
-------
1 1
K
H
ft
ft
«
EH
CO
s
W ft
EH "V
CO
O
i< in
« fa
o o
r_
t±4
IS
EH O
E5 i-l
fa" fa
EH &H
co fa-
fa" EH
*Z £
H fa"
EH
i4 CO
EH Sc
H
(^ ^H
< EH
CJ H
^~H
1-3 H
rfj 0
EH <
O fa
<;
fa
Ou
V-/
S HH
O EH
EH W
rf1 &.
C-j ^^
H
EH EH
CO H
W S5
£
H
1
in
Cd
•j
p_j
™
CO
o
•*-> §
O m
CJ (£
CO
PM
-P
^-Hg
(d -P fj
-P-H "ft
O P.I m
EH (d ^
CJ g
H
-P
C
ow5
Q^ O
u
o
^J
o
fa
^J
tn
o
CJ
-p
c
CO
o
ft
o
u
•p
in
O O
4J tn -H o -cj
WO CO
O O M 4J C
CJ C C ft (d
0) g Cd (d -H rH
g ft 4J d rH
P -H ex & (d
•H 3 co (d co -P
3 tXQ S -P W
o co cy 3 'O . c
w C M tn a) H
0) -H .p CO W
-P W ft W rH (d -P
O (d -H c rd fi O
CO W ft H CO O Q)
M r4 JH
•H 3 -H
Q ft Q
0 O O O O
CM CM CM CM CM
O O O O O
O O O O 0
O O O O O
n CM vo CM CM
CM r-
CJ CJ. ^ U CJ
ft ft <* ft ft
ft 1
ft I ft t T^ fl I fl 1
o o ° o o
o\» o\<> ^ o\<> o\°
CM rH ~r rH rH
rH
tn
^j
^
o tn
ft C
ft -H
3 rH
tn T3
C
C X!
id
Ti
- w C
C rH (d C
O cd o
-HOC &>-H
.p -H O C 4J
(d &-J "H -H (0
T3 -P -P -P rH
cooes
3 co co -H tn
O rH M Cd C
pLf pc3 W Pi H
o
o
o
in
o
H
EH
CO
O
O
&
ALLATIO
EH
CO
S
H
EH
U
fa"
H
Q
i_3
o^0 Q^0
iH t— i rH
CO
tn
c
c co
0 ft "
i ^
tn w co
•P -H
W > T3
O r4 rH
cj
•H -o tn
-P T3 C CO
cd c cd -P
• — i cd
rH C Ti
fd CT> O C ><
"4-> C -H Cd O
tn -H .p c
C rJ O ft CO
H CO 3 3 CT>
CO r4 1 C
4J C -P -P -H
O -H tn r< -P
CO tJi C cd C
rJ C O -P O
-H W CJ CO CJ
c
H
o
o
0
*•
in
vo
EH
CO
o
CJ
g;
o
H
STALLAT
H
EH
cj-
fa"
H
Q
H
jj
-------
H
H
I
tn
CQ
t
in
5-1
(d
rH
rH
O
TJ
fj,
CO
a\
rH
J>^
rH
£j
^
(d
assumed to total
anks and decanters.
of 5.1-cm diameter
in 4-> TJ
•H 0)
0) 4-)
in tr> O
g (d 3
(d 5-1 5-1
Q) O 4->
s-i 4-i in
in o
5-1 O
54 O
Q) (H rjj
•P C
•sol"5
0) 5-j X!
tn 3 tr>
(0 a1 C
£ Q) Q)
5-1 rH
Q)
> tn c
•H Q) -H
*w c
.Q
4J
(d 5-1 O
G 0 4->
•H ft
,Q (d T3
g > 0)
O g
O 3 •
• u a)
5-< g in ft
o (d-H
tW 0 ft
o in
C •« !3 0)
•H H 0)
ft Q) 4->
•H >i c in
fttH -H
Q) H O
i — I | * ^J*
3
4J O T3
•H 5-< X! 0)
T3 ft D ^5
T3 ft (d O
<< (d w tn
.
4->
in
o
0
-[.-1
c
0)
g
ft
•H
3
O1
0)
Q)
in
(d
ffl
II
o
w
ffl
o
.
_p
in
0
0
4J
c
d)
g
&
-r-|
3
O^
0)
^
a)
tn
cd
X!
0
S-i
3
II
O
W
p-l
•o
5-34
-------
EH
H
55
!=>
O
55
H
ft
H
EH
CO
IgM
<5 g
EH "W
CO
O
<; o
in
«
o fa
fa O
EH IS
55 O
fa i-3
S fa
CO PH
w w
H |S
w
< CO
H &
ft
CJ EH
H
< H
EH CJ
O <
EH fa
fa <
^ ^
O H
H EH
EH <
•rf1 (VI
P-*j H-4
S Pi
H EH
CO
fa
•
CM
i.
i
in
rvt
E"1
CU
O
-p §
cS ^
<4_l
0)
Pi
4J
, C
1 ' fli
rH CO £
(C J_) "
•»J-H In
0 ft$
EH CO ^
CJ ^
M
4-1 -—
Q) v_*
C
OcO
ft4->
g 01
O O
CJ CJ
O
O
fa
1)
•»•*
o
u
-p
c
f\\
\u
c
o
o
0
Jj
u
(Ti O O
H CM CM
O
o
o
*.
vo
o
^
o o o o
o o o o
o o o o
^3* CO ^ O
co in co co
CM
0) Q)
rH g
in en + +
E^ow
<«-2,^
^ * CO
H O
o o
4-1 01
4-1 03 4-1
J3 O 01
W 4-1 tP O O
4J 01 -H CJ
01 O 0) 4-1
00 ^ C C
CJ C M-l 0) O
4-1 O g -H
4-1 C -H T3 ft 4-1
C Q) 4J C -H CO
Q) g CO CO 3 rH
g ft 4-> O1 rH
O -H C X 0) CO
•H 3 CU CO 4->
3 WQ g 4J T» - 01
O 0) CT> 3 CU C
fa C in 01 01 H
Q) -H 4-1 O (0
4-> W ft 01 rH £3 4-1
O CO -H C (0 O »O
Q) OQ ft H CO k CU
^ 3 ^
•H ft -H
Q Q
O O O O O
CM CM CM CM CM
O O O O O
O O O O O
O O O O O
Q\ "^J* f^ ^* ^*
H
CJ CJ rn CJ U
W H tr 63 W
ft ft ft ft
fl 1
tt __i tit T^ fr j ti i
0 0 ° 0 0
0\° 0\° °~ 0\<> 0\°
CM H ~1, H H
iH
01
4J
O &>
ft C
ft -H
3 rH
c
T3 CO
C X!
CO
01 C
C H CO C
O CO O
-HOC tP-H
4-1 -H O C -P
CO k -H -H CO
T3 4J 4J 4-> rH
C O O C 3
3 CU CU -H 01
O rH !H CO C
fa W W ft H
o
o
o
^.
n
CM
CM
EH
CO
o
u
o
H
EH
*f£
EH
CO
H
EH
CJ
H
H
Q
J
rtj
EH
0
EH
o o o o o
CM CM CM CM CM
-
00000
o o o o o
o o o o o
H H rH ^ CM
\J >»J V— ' r \ r \
H W W X X
ft rV ft p4 (^
fl 1 ft I fl 1
O O O *Q *Q
o\« o\o o\o oV> ^
rH rH rH
CU
01
C
C CU
0 ft
•rH K^
03 01 0)
4-1 -H
01 > T3
O M rH
O CU CU &i
ft-H C
O 01 4J
•H T3 01
-P T3 C CU
CO C CO 4J
rH CO
rH C 'O
co ty O C >i
4J C -H CO O
01 -H 4-1 C
c rj o a cu
H 0) 3 3 &>
0) H 1 C
4J C 4-1 4J -H
O -H 01 M 4J
Q) &> C CO C
>H C O 4-> O
•H fa U CO O
'O
C
H
O
O
o
en
n
H
EH
CO
o
CJ
o
H
|
1
CO
55
H
EH
CJ
H
Q
55
H
J
-------
EH
H
S5
D
13
*25
H
pL|
P[
H
EH
CO
3 g
W ft
EH in c
•H Q) -H
(II P*
*fl M
X3
-P
fd rH O
C 0 -P
•H ft
•Q fd T3
g ^* Q)
0 g
O 3 •
. m CD
!H g in ft
o rd -H
•H -H >^ c in
O ftrH-H
TJ (!) rH O
•H 4J <*
C\ (d fd 4->
CO C g C 0)
CT> O -H 0) rH
H -H X > 3
4J O tj
>i -H j-i x; Q)
rH Ti ft 0 X!
3 TS ft fd O
(d A
•
4J
in
0
o
4J 4J
in c
O Q)
0 g
ft
4J -H
(D CT1
g Q)
ft
-H T!
3 Q)
c? in
Q) fd
X!
Q) U
in JH
fd 3
CQ f\|
u u
u u
W W
CQ CU
O 13
5-36
-------
O
53
H
PH
Pn
H
EH g
W ft
M
<
C
O
• r-j
"(0
3 e
ll
^^ f1
O
cj
JH
O
•P
O
(0
fe
j_)
tn
o
-P
C
0)
C
o
Q,
E
O
cj
-P
Ifl
o
cj
rH H CM O O O O
CM CM CM CM CM CM CM
° 0 00 00
O Q O O O O O
O O O O O O
** CM" "" - " - -
^ in vo rH r-- t-~
X) "^
^| X^
JC ^4 S W W
So ^ ^
" O «• "^ "^
0§§ 0 0
° - ° ^ 15
*• S i-T
^ *
^
o
rH XI
c^ O (0
r-l XI rH
X3 0 j. «3 ,,
g: t7>o 2 rH 2 •
^ g 0 ^ ^4J
^^ ^*>- ^ Q * O 5
lO * CO E
^^ f*^ l^ J (|—j ,
o ^^ ° 'c/> o
O o\»
•cn- in o\<>
H 0
0
H
<
n
•H
cs
T3
cd
rH t3
0 C
CO X? <0
•P (d
CQ >1 rH C
O -P O
cj -H tri-H 1/3
OEM C W 0) rH
r-H -rH 03 0) -H -H O (d
(d Wr-lCD-P -P> C-H
£f 0) -P -P fd fd rH (d r-i JH
C -HOWS MO) COO)
C-PQ) >]i p, e\\ o _p
rt3 -H i — i JH PI ^ 4J td fd
r-HW OOW CJg
4J -H X( -H
O 4-> ns fd
Q) t> J a
•H
Q
O
O
o
CO
•*
EH
o
u
1-H
£j
55
s
EH
U
p^
H
Q
d
J EH
O
EH
O O O O O
CM CM CM CM CM
! lit!
n ^T ^T ,C >
JH
0 w u
n3 id 9. H H b*
rH -H tl U CJ •>(
LJ "-^ C-l C-l
rH i~ • u1 C^ t/-\
rH 0) ^ 2
rH -P (^ OH <4H M
10 g O
UH ,0 o\° o\» ,0
O T3 ^ H CM Q
o\<> fd H
O
W
Q)
^
tj
to fd
tfl O >i
0 rH
a w a) >
rH X -H 0
rd fd -P o
3 -P 03 Q)
C Q) r-l rH
C 73 ^i O -P
< fd -P C U i-H
0) rH fd -H fd
-P x! (U JH C -P
O JH ft S -H -H
a) a) o w g a
r-l > rH C 73 fd
•H O C4 H < O
73
H
o
0
o
in
EH
CO
o
u
Hi
-------
0
SB
H
0*
H
EJ
fcH rt
CO "
g
w S
EH
W (^
•< °
/v &
**"" f*4
03
o i3
o Hj
j JS ^
"^ EH Q
S CO Q
Ps
J >H igj
g§8
^o
fe p4
<
o o
H s
jH H
M* r .
^^ u*
*• fVI
S
f^ EH
PH
EH
«i
H :D
1
in
TO
r^
IS
Q)
O
°l
0)
Pi
^
rd'Jj _
^ n ^
f^ p^
rtj U
C
O
1-1 4J
S&
C g
< w
o
CJ
0
-p
o
frt
l*J
-p
0
CJ
Cost Component
q_i
o
0
o
•fc
CM
•
RECOVERY CREDIT
o
o
rH
CM
H
CJ
PS
1
0
H
EH
CJ
Q
EH
TOTAL ANNUAL COST
,-j,
o
0 ° VD
o
•CO-
o
o
o
o" >,
o \
CO H -P
M >i tfl
JH >. \ Q)
aJuly 1989 dollars.
b24 kWhr/day, 300 days/y
C9,700 kg/ day, 300 days/
d261,000 £/day, 300 days
OTCI _ Total capital inv
.
cu
rH
rd
>
&i
C
-H
-P
rd
Q)
XI
CT>
W
j!>
w
o
o
o
co
CM
rH
CU
-p
(d
•H
X
o
^
ft
rd
C!
^Recovery credit based o
CO
o
u
cu
1
1
.p
in
o
CJ
id
g
C
<
-P
O
cu
>H
-H
^
fj
H
rH
(d
O
EH
+
•P
O
U
rH
rd
g
rf!
•^s
-p
O
0)
M
•H
Q
rH
9TDAC + TIAC - RC = Tota
Credit
IS
o
rH
)_!
CU
ft
4J
Ul
O
CJ
hTAC/Flow = Total Annual
5-38
-------
CD
55
H
ft
ft
H .
EH fi
W ft
g
(^ O
W 0
EH m
w
fa
H
< EH
EH H
0 i-3
EH H
U
fa rtj
O fa
S3 ~
g u
5 C-)
"^3
G
o
*™H i)
"^ ft
< g
O
u
M
O
4-1
O
(0
fa
4-)
in
o
0
4->
C
0)
o
ft
K
o
u
4-J
U
o
o
H H CM O O O O
^ W t i -s^^^ *W It
^^* ^* C5 , Q O ^ Q
gCOr-1 (0
W >1 rH C
O 4-J 0
U -H JJi-H W
O g M C Cfl 0) H-
rH -H (0 (U -H -H O J g
•H
Q
o
o
O
*.
O
en
CM
EH
O
U
£5
^
s
<
EH
U
H
H
Q
<
EH
O
EH
o o o o o
CM CM CM CM CM
O O O O O
O ' O O O O
O O O O O
•. «. «. ^ ^
n co co in H
H HO
H
m
(d (0 cl EH EH ,«
1 t™1 ^ J
*i | "n i
_ M M Q, 'W 'H
O Q) /-» O O rtj.
\o «^ "^
Q (0 (0 \0 o\° o\° ,0
i-% (~~f S \ H CM xX
H
0)
&*
u a
4J X!
W 0 >i
O M
O W 0) 0)
Q) ^ ^
rH X -H 0
(d (0 4-1 O
3 4J (0 a)
C 0) M M
C t3 >i 0 4-1
M C T3 (0
•H O ft H
-------
0
53
H
O<
CM
I,
co a
w o
EH in
CO
fa
rfj O
fa
s
fa
w
"33 fa" P
e fr* M
fe W Q
S3 a
EH O
H O
O
O
H O
EH S5
rf H
ijj4 r.
H
U
H
EH
U
P
CO
U
•co-
CO
00
VO
OJ
EH
CO
EH P
CO W
O W
CJ fa
IQ
0)
(1)
CQ
&
73
(0
T!
cn
CO
CTl
(C
T!
^ vo
VO
in
o\
O
(N
ft
(0
o
EH
H
U
EH
0)
id
(U
o
CO
CN
0)
-P
(0
ft
ft
(0
W
(d
;Q
T!
0)
^
O
0)
o
0)
o
Q)
O
-------
55 2
I 1
CO O
m +
CM
T—
O
o
O
ci
q
o
I
a>
8
d
I
1
o
O
o
00
I
Q)
-P
0)
Q)
Q)
-P
(0
0)
-P
in
(0 •
5 -P
•H
01 C
3 3
tn
^ IT
0) C
01 M
O -P
O 01
C (0
•H Q)
-P -P
(0 ui
^1
0) M
-P
•H
C
in
0)
8
ci
5-41
-------
in
H
I
in
PQ
tn
w
0)
0)
w s
o
c
&
Q)
CO
O
JJ'H
II
'm (0
,2 0)
3 H
oooooooooooooooooo
oooooooooooooooooo
ooooonoo-^ooooor^ooin
O O O O O
O O O O H
O O VO VO cH
O O] CN
H H
inocoHooooininofM
CTlCN OOOi-iTj" ^H
^o c^ ro r** o^ 03
o in
H
H
rH O
CM
CO
LD OVOCN O
OOfOOfO COCO OHCOCTl Lf)
OOOOCNVOOCOlOOOHOCOt^COOO
OOOOO'3'OOt--OOOOH-*O
H H
IT)
vo
H o> en
O CO CT> H CO
o o r-» o co
VO VO
CO CTl H CM CO VO
OCO<7lOCOinHVOOCOi-ICN
ooooovooor-oooqocMHot^r-
H CM IT)
oooooooooooooooooo
oooooooooooooooooo
CVJCVJCMCNCNJCMCNCXJCNICOCOCOCOCOCOCOCOCO
oooooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
HHHHHHHHrHCOCOCOCOCOCOCOCOCO
in vo
5-42
-------
o
8
(O
Q.
73
O
CD
I
5
I
0)
-p
(0
M
0)
0)
(1)
-p
(0
01
^-x m
0)
0)
ta
0)
CO
0)
(0
x
(U
-H
4J
O
0).
0)
-p
u
O
O
I
If)
0)
(0
00
CO
CO
sseu8A!;oajj3
5-43
-------
5.5 REFERENCES
1. U. S. Environmental Protection Agency. Office of Air
Quality Planning and Standards and Office of Research and
Development. Control Technology Center, Industrial
Wastewater Volatile Organic Compound Emissions -
Background Information for BACT/LAER Determinations.
EPA 450/3-90-004. Research Triangle Park, NC. January
1990.
2. Code of Federal Regulations, Title 40, Part 60,
Subpart W, Sections 60.481 and 60.489. Standards of
Performance for Equipment Leaks of VOC in the Synthetic
Organic Chemicals Manufacturing Industry. Washington,
DC. U.S. Government Printing Office. July 1, 1990.
3. Memorandum from Zukor, C., Radian Corporation, to
Lassiter, P., EPA/CPB. January 27, 1992. Development of
national impacts from responses to the March 1990 Section
114 wastewater questionnaire—A summary of the Section
114 database.
4. Memorandum from Zukor, C., Radian Corporation to
Lassiter, P., EPA/CPB. January 27, 1992. Development of
model wastewater streams for the HON.
5. Memorandum from Zukor, C., Radian Corporation, to
Lassiter, P., EPA/CPB. January 20, 1992. Approach for
estimating uncontrolled emissions of hazardous air
pollutants from wastewater streams in the HON.
6. Memorandum from Watkins, S., Radian Corporation, to
Lassiter, P., EPA/CPB. January 31, 1992. Model
wastewater stream selection to present representative
impacts for the HON.
7. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Fossil Fuel Fired
Industrial Boilers - Background Information, Volume I:
Chapters 1-9. Draft EIS. EPA-450/3-82-006a. Research
Triangle Park, NC. p. 3-12 to 18.
8. U.S. Department of Energy. Electric Power Quarterly,
April to June 1984. DOE/EIA-0397C84/2Q. Energy
Information Administration. Washington, D.C. October
1984. pp. 19, 20.
9. U.S. Environmental Protection Agency. Office of Air
Quality Planning and Standards. Compilation of Air
Pollutant Emission Factors. Volume I: Stationary Point-
and Area Sources. EPA/AP-42. Research Triangle Park,
NC. September 1985, and Supplement A, October 1986.
pp. 1.3-9, 1.3-4.
10. Ref. 9, pp. 1.1-5, 1.1-6.
5-45
-------
11. Ref. 9, pp. 1.1-2, 1.3-2, 1.4-2.
12. Corripio, A. B., K. S. Chrien, and L. B. Evans. Estimate
Costs of Heat Exchangers and Storage Tanks via
Correlations. Chem. Eng. January 25, 1982. p. 145.
13. Peters, M. S., and K. D. Timmerhaus. Plant Design and
Economics for Chemical Engineers. Third Edition. New
York, McGraw-Hill Book Company. 1980. pp. 768 - 773.
14. Cqrripip, A. ;B., A. Mulet, and L. B. Evans. Estimate
Costs of Distillation and Absorption Towers via
Correlations. Chem. Eng. December 28, 1981. p. 180.
15. Hall, R. S., W. M. Vatavuk, jT. Matley. Estimating
Process Equipment Costs. Chem. Eng. November 21, 1988.
pp. 66-75.
16. Ref. 13, p. 572, Figure 13-58.
17. Teleeon. Gitelman, A., Research Triangle Institute, with
Hoyt Corporation. September 8, 1986. Cost of flame
arrestors.
18. U.S. Environmental Protection Agency. Office of Air
Quality Planning and Standards. OAQPS Control Cost
Manual. Fourth Edition. EPA-450/3-90-006. Research
Triangle Park, NC. January 1990. p. 2-5 to 2-8.
19. Richardson Engineering Services, Inc. Richardson Process
Plant Construction Estimation Standards: Mechanical and
Electrical. Volume 3. Mesa, AZ. 1988. pp 15-40.
20. Vatavuk, W. M., and R. B. Neveril. Part II: Factors for
Estimating Capital and Operating Costs. Chem. Eng.
November 3, 1980. pp. 157-162.
21. Memorandum from Peterson, P., Research Triangle
Institute, to Thorneloe, S., EPA/CPB. January 18, 1988.
Basis for steam stripping organic removal efficiency and
cost estimates used for the source assessment model (SAM)
analysis.
22. Ref. 14, p. 4-29.
5-46
-------
-------
6.0 EQUIPMENT LEAKS MODELS
Standard types of equipment used at facilities within the
SOCMI have the potential to emit VOC's, many of which are
organic HAP's. Emissions from equipment leaks are typically
released directly to the atmosphere. The techniques used to
control equipment leak emissions can be classified as either
work practices or equipment design specifications.
This section discusses potential equipment leak emission
sources, models used to represent equipment leaks, and impacts
of the control technologies on SOCMI process units.
Section 6.1 discusses different types of equipment that have
emissions. Section 6.2 discusses development of model units.
Section 6.3 presents environmental and energy impacts of
controlling equipment leaks, and the cost impacts of the
control technologies are presented in Section 6.4.
6.1 EMISSION SOURCE DESCRIPTION
This section provides a brief description of the
potential equipment leak emission sources that are typically
found in the SOCMI. More detailed descriptions of the
potential equipment leak emission sources along with various
control options to reduce emissions from these sources are
presented in the HON BID Volume IB.
The focus of this study is VOHAP emissions associated
with equipment leaks that result when process fluid (either
liquid or gaseous) is lost or released from various types of
equipment. The following potential sources of equipment leak
emissions are considered in this chapter: pumps, compressors,
process valves, pressure relief devices, open-ended valves or
lines, sampling connections, flanges and other connectors,
agitators, product accumulator vessels, and instrumentation
6-1
-------
systems. Leaks in these sources are random occurrences that
cannot be predicted on an individual component basis, and leak
occurrence is generally independent of temperature, pressure,
and other process variables. The population and distribution
of agitators and product accumulator vessels in the SOCMI have
not been characterized, and there are insufficient emissions
data for these equipment to permit adequate estimation of
industry-wide impacts. Instrumentation systems consist
primarily of valves and connectors, which are included in two
other equipment categories. The impacts for instrumentation
systems, therefore, are included with the impacts for these
other equipment types.
6.1.1 Pumps
Pumps are used extensively in the SOCMI for the movement
of organic liquids. The centrifugal pump is the most widely
used pump design in the SOCMI; however, other types, such as
positive-displacement pumps, are also used. Chemicals
transferred by pump can leak at the point of contact between
the moving shaft and stationary casing. Consequently, all
pumps except the seal-less type, such as canned-motor,
magnetic drive, and diaphragm pumps, require a seal at the
point where the shaft penetrates the housing in order to
isolate the pumped fluid from the atmosphere.
6.1.2 Compressors
Gas compressors used in the SOCMI can be driven by rotary
or reciprocating shafts. Seals must be used between the shaft
and housing to isolate the process gas from the atmosphere.
As with pumps, these seals can be a source of equipment leak
emissions from compressors.
There are several different types of shaft seals for
compressors including labyrinth, restrictive carbon rings,
mechanical contact, liquid film, and packed seals. All of
these seal types restrict leakage, although none of them
completely eliminate leakage.
6-2
-------
6.1.3 Process Valves
One of the most common pieces of process equipment in
organic chemical process units is the valve. Valves are
available in many designs, but most contain a valve stem that
adjusts the plug, thus restricting or allowing fluid flow.
Typically, the stem is sealed by a packing gland to prevent
leakage of process fluid to the atmosphere. However,
emissions from valves occur at the stem or gland area of the
valve body when the packing or O-ring in the valve fails.
6.1.4 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 is the pressure relief valve.
Typically, relief valves are spring-loaded and designed to
open when the internal pressure exceeds a set pressure,
allowing the release of vapors or liquids until the internal
pressure is reduced back to the set operating level. When the
normal 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 leakage through this seal makes the pressures
relief valve a potential source of emissions. Two potential
causes of leakage from relief valves are "simmering or
popping," a condition caused by the system pressure being
close to the set pressure of the valve, and improper reseating
of the 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 emissions as long as the integrity of the
disk is maintained. The rupture disk must be replaced after
each pressure relief episode to restore the process to the
condition of no emissions. Although rupture disks can be used
alone, they are sometimes installed upstream of a pressure
relief device to prevent emissions through the relief valve
seat.
6-3
-------
6.1.5 Open-Ended Valves or Lines
Some valves are installed in a system so that they
function with the downstream line open to the atmosphere.
Examples are purge valves, drain valves, and vent valves. A
faulty valve seat or incompletely closed valve can result in
leakage through the valve and emissions to the atmosphere.
6.1.6 Sampling Connections
The operation of a process unit is checked periodically
by routine analyses of feedstocks, intermediates, and
products. To obtain representative samples for these
analyses, sampling lines must 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 resulting
in emissions to the atmosphere.
6.1.7 Connectors
Flanges, threaded fittings, and other fittings used to
join sections of piping and equipment are connectors. They
are used wherever pipe or other equipment such as vessels,
pumps, valves, and heat exchangers may require isolation or
removal. Normally, flanges are used for pipe diameters of
50 mm or greater and are classified by pressure and face type.
Connectors may become emission sources when leakage
occurs due to improperly chosen gaskets or poor assembly. A
common cause of connector leakage is thermal stress that
piping or connectors in some services undergo which results in
the deformation of the seal between the connector parts.3
Improper installation of the connectors can also result in
equipment leak emissions.
6.1.8 Agitators
Agitators are used in the SOCMI to stir or blend
chemicals. As with pumps and compressors, emissions from
agitators may occur at the point where a moving shaft
penetrates a stationary casing. Emissions from this source
may be reduced by improving the seal at the junction of the
shaft and casing. Four seal arrangements are commonly used
with agitators: packed seals, mechanical seals, hydraulic
4
seals, and lip seals.
6-4
-------
6.1.9 Product Accumulator Vessels
Product accumulator vessels are small, primarily fixed
roof storage tanks designed to regulate material flow through
a process. They include overheads and also bottoms receiver
vessels used with fractionation columns and product separator
vessels used in series with reactor vessels to separate
reaction products. Emissions occur when gases are vented to
the atmosphere either directly or through a blowdown drum or
vacuum system.
6.1.10 Instrumentation Systems
An instrumentation system is a group of equipment
components used to condition and convey a sample of the
process fluid to analyzers and instruments for the purpose of
determining process operating conditions (e.g., composition,
pressure, and flow). Valves and connectors are the
predominant types of equipment used in instrumentation
systems, although other equipment may be included. Normally,
instrumentation system equipment components are less than
20 mm in size.
6.2 MODEL EMISSION SOURCE DEVELOPMENT
This section describes the development of model process
units that were used to assess the environmental and cost
impacts of the negotiated regulation for equipment leaks.5 In
general, emission estimates are based on the number of
equipment components in a process, with most emissions
attributed to components that are leaking. This approach had
been used to assess impacts of proposed equipment leak
regulations, including the SOCMI NSPS and the benzene NESHAP,
as described in the Fugitive Emissions Additional Information
Document (Fugitive Emissions AID)6 and in the Background
Information Document for Benzene Fugitive Emissions,7
respectively.
Each model process unit consists of a combination of
pieces of equipment representing a typical SOCMI process unit.
The equipment counts are the fundamental elements that define
the model units. In addition, the models represent the
6-5
-------
differences in existing regulations controlling emissions from
SOCMI units.
Emissions sources considered in this section include
heavy-liquid and light-liquid pumps; compressors;; gas, light-
liquid, and heavy-liquid valves; pressure relief valves; open-
ended lines; sampling connections; and connectors.
Emission factors were developed for each component type
and level of emission control. Model unit emissions and
emission reductions were then calculated by multiplying the
emission factors by the number of components for each model
process unit.
6.2.1 Data Gathering
The approach for developing model units for this analysis
is similar to that used in the Fugitive Emissions AID.6
However, component counts were based on different data
including EPA's study of 24 process units in the SOCMI
(24-Unit Study)8 and studies of equipment leak frequencies and
emissions at 32 butadiene, ethylene oxide, and phosgene
production units.9 These newer data reflect more current
design of chemical production units. Additionally, the data
from the 24-Unit Study8 differentiate between equipment in
light-liquid versus heavy-liquid service, while the older data
used in the Fugitive Emissions AID did not.
Baseline emission factors for components assessed in the
model analysis were obtained from the Fugitive Emissions
AID.10'11 Controlled emission factors were developed from
information provided in the negotiated regulation and are
based on data from a series of tests conducted by the EPA and
Chemical Manufacturers Association (CMA) butadiene, ethylene
oxide, and phosgene panels.9 Control costs used in this
assessment were based on information in the Fugitive Emissions
AID,12 updated with information supplied by vendors.
6.2.2. Model Development
Model units were developed to reflect the differences in
process unit complexity and level of existing regulatory
control at process units in the SOCMI. Three levels of
6-6
-------
complexity were represented by different equipment counts.
Two levels of existing control were also represented. These
two parameters were combined to form a matrix of six model
units. Equipment counts and levels of control are discussed
below.
6.2.2.1. Number of Equipment Components. For valves and
pumps, model unit equipment counts were developed for
different classes of volatility.
Data collected in the late 1970s from petroleum
refineries indicate that emission rates of sources decrease as
the vapor pressure (volatility) of the process fluid
decreases. Three classes of volatility were established based
on the petroleum refinery data; these include gas/vapor
service, light-liquid service, and heavy-liquid service.13
The split between light and heavy liquids for the refinery
data is between naphtha and kerosene. Because similar 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. Available data aTso indicate that
equipment leak emissions are proportional to the number of
potential sources, but are not necessarily related to process
capacity, process throughput, component age, operating
temperature, or operating pressure. Therefore, SOCMI model
units defined for this analysis represent different levels of
complexity (number of sources) rather than different unit
sizes. Three levels of complexity were assessed: simple,
medium, and complex.
The model units were developed from a data base compiled
from the SOCMI 24-Unit Study and the Ethylene Oxide/
Butadiene/Phosgene Study9 and included equipment counts from
56 SOCMI units. Equipment counts for these units were used to
develop model unit equipment counts for light- and heavy-
liquid pumps, gas/vapor and light- and heavy-liquid valves,
open-ended lines, pressure relief valves, and compressors.
Frequency distributions of the equipment counts were grouped
6-7
-------
into three population ranges, which represented simple,
medium, and complex process units. Simple process units had
the fewest components and complex process units had the most
components. The median number of components within each
population range was selected as the model unit equipment
count.
Equipment counts for connectors and sample connections
were reported infrequently or incompletely. Therefore,
connector counts were estimated using a connector-to-valve
ratio (1.6:1) determined from a study of eight SOCMI process
units.15 The estimated number of sampling connections in each
model unit was based on data showing that 25 percent of open-
ended lines are used for sampling.
Table 6-1 presents the model unit equipment counts for
simple, medium, and complex model process units. These model
unit equipment counts are assumed to represent the range of
emission source populations that may exist in SOCMI process
units.
6.2.2.2. Existing Level of Control. In addition to the
complexity of a unit, a major parameter that impacts equipment
leak emissions from a process unit is the level of control
within the process unit. For this analysis, the level of
control existing in a process unit determines how the baseline
emissions are evaluated. Baseline emissions are the emissions
before implementation of the controls required by the
negotiated regulation.
In 1984 EPA published a CTG document on control of VOC
emissions from equipment in the synthetic organic chemical and
polymer manufacturing industries.17 Several States relied on
the CTG when they adopted SIPs for areas that have not
attained National Ambient Air Quality Standards,. Compliance
with the SIPs has helped reduce equipment leak emissions from
many process units. Other facilities, however, are not
subject to equipment leak VOC regulations and have not taken
formal measures to control VOHAP or VOC emissions from
equipment.
6-8
-------
TABLE 6-1. MODEL UNIT EQUIPMENT COUNTS
Equipment Counts
Equipment Component
Pump seals
Light-liquid service
Heavy-liquid service
Compressor seals
Valves
Vapor service
Light-liquid service
Heavy-liquid service
Pressure relief devices
Open-ended valves
Sampling connections
Connectors
Simple
15
0
0
77
380
0
2
33
8
731
Medium
40
5
2
414
1,179
71
45
141
35
2,662
Complex
56
36
8
1,379
1,980
1,272
76
424
106
7,410
6-9
-------
Two levels of baseline control are assumed for the model
unit analysis: (1) no controls required by any regulations,
and (2) control to the level discussed in the SOCMI equipment
leaks CTG.18 The model units with no required controls are -
assumed to have emission factors equivalent to the average
emission factors presented in the Fugitive Emisssions AID10 for
all equipment types except pressure relief devices and open-
ended lines. Emissions from pressure relief devices and open-
ended lines are commonly controlled in process xmits, and
therefore, 75 percent of the pressure relief devices and
100 percent of the open-ended lines in the uncontrolled model
units are assumed to be controlled to the level discussed in
the CTG.18
The techniques used to control equipment leak emissions
may be classified as either work practices or equipment design
specifications. Work practices include leak detection and
repair (LDAR) methods to identify and control equipment
components that are larger sources of emissions,. Equipment
design specifications include use of improved valve packing,
flange gaskets, and pump and compressor seals as well as use
of control equipment such as caps or plugs for open-ended
lines, closed-vent systems for pressure relief valves, and
closed-loop sampling systems.
Many process units have employed a combination of
equipment control techniques as part of an emission reduction
program. Motivation for development of such programs has
included regulatory compliance, voluntary participation in
efforts to reduce airborne HAP emissions, and practical
concerns for protection of workers from toxic chemical
exposure or minimization of losses of expensive chemicals.
Unit-specific emission reduction programs vary in
stringency. However, most programs resemble the RACT
17
procedures outlined in the CTG. These procedures include
capping of open-ended lines and quarterly LDAR of pumps in
light-liquid service, valves in gas/vapor and light-liquid
service, compressor seals, and pressure relief valves. The
6-10
-------
control efficiencies estimated in the CTG for these procedures
are 33 percent for light-liquid pumps, 64 percent for gas
valves, 44 percent for light-liquid valves, 44 percent for
safety relief valves, 100 percent for open-ended lines, and
33 percent for compressor seals.18 This "CTG level" of control
was assumed for the model units with some existing controls in
place.
6.2.2.3 Model Characteristics. The six model process
units developed are designated by the letters A through F.
Model units A, B, and C are uncontrolled and have equipment
counts representing simple, medium, and complex process units,
respectively. Model units D,, E, and F are controlled and
represent simple, medium, and complex process units,
respectively. The equipment counts and level of existing
control for the six model units are listed in Table 6-2.
6.3 ENVIRONMENTAL AND ENERGY IMPACTS OF CONTROLLING EMISSIONS
FROM EQUIPMENT LEAKS
The environmental impacts resulting from the
implementation of the HON equipment leaks standard on the
model units are discussed in this section. Impacts have been
grouped into primary and secondary impacts. Primary
environmental impacts of the regulation occur from the
reduction of HAP and VOC emissions. Secondary impacts include
changes in water quality, solid wastes, and energy use.
6.3.1 Primary Environmental Impacts
Baseline emissions, controlled emissions, and emission
reductions are estimated for each model unit based on assigned
equipment counts, level of existing baseline regulatory
control, and maximum achievable control technology (MACT)
emission factors. The "MACT controlled emissions" are
emissions from components controlled to the level stated in
the notice of agreement on the negotiated regulation.5 To
estimate national impacts, each affected facility in the
United States is assigned a model. This assignment of models
to the population of affected SOCMI product processes is
discussed in BID Volume 1A, Chapter 4.
6-11
-------
en
S
At
C~i
H
s
1
•
1
1
vo
ffl
i?
C_I
cH
Counts
g
8
cii
>iH
0*
ta
cd
,ti >
U U
C C
0) 0)
B C
cu o
~3 I4
cr1 o
W O
VD VO CO O*» O O-J ^O
in co r^- oo t*^. r*-
OO CT\ 0^4
rH r-l i-l
o m CM i— i in
«ct" i~H 1^^ P"^ *«d°
I-H ,n
3 3 rH
& cr k cr cr a)
0] -iH -iH !-4 0) >iH -rl )-l
,_! ,_| d) ^_( 4) o W r— 1 0) i— Id)
nj cj o w i o i o a)
(1) 4J -,H !>-, -r-l W W -U -iH t>, -r-l !-l W
w ,£>!>> 0) w o,C>F>> 30)
bO M cd M ^i 0) P-i CiO M cd J-i u] O
§ o cd Vi Q)
PH u j> . PM "d
CM O
C
•H
T3
0)
•rH
<±4
•H
0
0)
A
U)
r-l
fl)
fl)
rH
0)
4-)
o
-p
Tf
fl)
rH
rH
O
•P
^
O
U
II
T5
fl)
rH
r-l
O
i-l
-p
o
o
o
EH
O
(0
•
fl)
c
-H
i — 1
0)
(0
XI
4J
(0
•o
0)
rH
O
J^
-p
G
O
O
M)
EH
O
0)
XI
0
TJ
0)
£3
CO
U)
(0
fl)
(0
fl)
rH
as
fl)
•H
H
0)
i^
^^
^i
-P
fl)
(0
U)
o
•p
f"j
Q)
O
0)
p.
0)
^
•H
i
event
w
XI
,
G
•H
0)
to
(0
XI
[^
rt
-d
0)
rH
rH
o
.p
c
o
o
o
EH
U
fl)
XI
o
-p
sumed
w
ti
a)
rH
(0
U)
fl)
C
•H
rH
TJ
fl)
rd
(3
fl)
1
G
0)
ft
O
O
-p
G
fl)
O
fl)
a
-d
d)
rH
T3
c
1
0)
c
o
o
6-12
-------
It should be noted that a default approach was taken for
facilities with low capacities (under 1,000 Mg/yr).19 A model
unit was not assigned to these facilities because baseline
emissions would have been an unrealistically high percentage
of capacity. In fact, for some of these small facilities, the
capacity was below the baseline emissions of model units A and
E, which have the lowest equipment counts. For these small
process facilities, baseline emissions were calculated, as a
percentage of capacity, with emission reductions and control
costs assumed to be zero.19
Model unit emission reductions were calculated as the
difference between baseline emissions and emissions remaining
after the implementation of the standard.
Baseline and MACT emission factors for the equipment
types are presented in Table 6-3. The factors can be used to
estimate VOC emissions from these typical sources. The
baseline factors represent (1) average uncontrolled SOCMI
emission factors,10 and (2) emission factors associated with
the controls specified in the SOCMI equipment leaks CTG.18 The
MACT emission factors represent emissions judged typical of
the control levels specified in the^negotiated regulation.5
Because the RON regulates HAP emissions rather than total
VOC, estimates of HAP emissions and emission reductions were
estimated as part of this analysis. These emissions were
estimated using an industry-wide ratio of HAP usage to VOC
usage in affected process units. Using stoichiometric
information available for 135 processes, this ratio was
determined to be 0.78. This ratio does not apply to any
specific HAP or process unit.
6.3.1.1 Maximum Achievable Control Technology Emissions
from Equipment Controlled With Leak Detection and Repair. The
negotiated regulation5 requires an LDAR program to reduce
emissions from light-liquid pumps, gas and light-liquid
valves, and connectors. The level of control attributable to
the negotiated regulation is considered MACT for equipment
leaks. To estimate MACT emissions from these pieces of
6-13
-------
TABLE 6-3. SUMMARY OF EQUIPMENT LEAK VOC EMISSION FACTORS
Baseline Emission Factors
MACT
Controlled
Emission
Equipment
Pump seals
Compressor
seals
Valves
Service Uncontrolled
(kg/hr)
Light liquid
Heavy liquid
Gas/vapor
Gas
Light liquid
Heavy liquid
0.0494
0.0214
0.228
0.0056
0.0071
0.00023
CTG
Controlled
(kg/hr)
0.0331
0.0214
0.153
0.002
0.004
0.00023
Factors
(kg/hr)
0.00248
0.0214
0.0223
0.0001216
0.000717
0.00023
Pressure
relief
devices
Open-ended
lines
Sampling
connections
Connectors
Gas /vapor
All
All
All
0.104
0.0017
0.015
0.00083
0.0582
0.0
0.015
0.00083
0.0
0.0
0.0
0.000345
6-14
-------
equipment, VOC concentration data from equipment at 19
butadiene and ethylene oxide units were analyzed.9
The negotiated regulation5 provides different definitions
of "leak" for each equipment type. If equipment has a
monitored value over a specified concentration then it is
considered to be "leaking." The negotiated regulation also
sets base performance levels specifying the maximum percentage
of equipment at any given time that can be leaking. For
example, a unit must demonstrate that less than 2 percent of
all valves have monitored concentrations (also known as
screening values) above the leak definition of 500 ppmv. The
leak definitions and base performance levels required by the
negotiated regulation are presented in Table 6-4.
The procedures used to calculate average MACT emission
rates given the level of control stipulated in the negotiated
regulation are outlined below. First, for each component
type, the relationship between average unit emission rates and
the percent of components at the unit that are above the leak
definition was determined using a regression analysis.20 From
this, two average emission rates were calculated: (1) the
average emission rate for a unit operating at the performance
levels specified in the negotiated regulation (i.e., 2 percent
of all valves leaking at or above a leak' definition of
500 ppmv), and (2) the average emission rate for a unit with
no equipment leaking above the specified leak definition.
For this analysis the emission rate assumed to be
associated with MACT was calculated as the average of these
two rates. The rationale for using this average was that over
the course of the LDAR cycle, maximum emissions would exist
immediately before repair, while emissions after repair would
reflect units with no equipment over the leak definition.
6.3.1.2 Maximum Available Control Technology Emissions
Rates for Other Equipment Types. MACT emission rates for
compressors, pressure relief devices, open-ended lines, and
sample connections are based on design specifications for the
equipment.
6-15
-------
TABLE 6-4. LEAK DEFINITIONS AND BASE PERFORMANCE LEVELS
FOR PUMPS, VALVES, AND CONNECTORS
Equipment
Leak Definition
(ppmv)
Performance
Level
Pumps
Valves
Connectors
1,000
500
500
10.0%
2.0%
0.5%
6-16
-------
6.3.1.2.1 Compressor seals. The negotiated regulation5
requires equipment specifications for compressors that
includes a seal system with a heavy-liquid or non-VOC barrier
fluid that prevents leakage of the process fluid to the
atmosphere. Estimates of the control efficiency of this type
of system were presented during development of the SOCMI
NSPS. Basically, the overall control efficiency assumes
95 percent capture of VOC emissions to a closed-vent system
and 95 percent reduction of the captured emissions in a
control device.
The MACT-level emissions from compressor seals are equal
to the VOC emissions not captured in the vent system plus the
VOC emissions remaining after the control device. The MACT
emission factors for compressors is calculated as follows:
EMACT =
* o.os) +
* 0.95 * o.os)
where:
EMACT
EUN
= Emission rate under MACT control
= Uncontrolled emission rate
= 0.228 kg/hr/source
EMACT = (0.228 * 0.05) + (0.228 * 0.95 * o.os)
= 0.0223 kg/hr/source.
6.3.1.2.2 Pressure relief devices. The negotiated
regulation5 requires that pressure relief devices be operated
with an instrument reading of less than 500 ppmv above
background at all times. The combination of rupture disk and
relief device is nearly 100 percent effective in controlling
emissions from relief devices, provided the integrity of the
disk is maintained.
6.3.1.2.3 Open-ended lines. The negotiated regulation6
requires that open-ended lines be equipped with a plug, blind
flange, or second valve to prevent emissions through the open
end. This approach to control is assumed to be 100 percent
effective in controlling emissions from open-ended lines.
6.3.1.2.4 Sample connections. The negotiated
regulation requires that sample connection systems be closed-
loop, thereby preventing any emissions of the process fluid to
6-17
-------
the atmosphere. Closed-loop or in situ samplingr systems are
assumed to be 100 percent effective in controlling emissions.
6.3.1.2.5 Equipment in heavy liquid service. The
negotiated regulation5 does not include work practice or
equipment design requirements for equipment in heavy liquid
service. These pieces of equipment must be monitored only if
leakage is suspected by visual, audible, or olfactory methods.
There are no specific performance standards that must be met
for equipment in heavy-liquid service and, therefore,
estimates of emission reductions due to implementation of the
negotiated regulation5 cannot be calculated for this type of
equipment.
6.3.1.3 Model Volatile Organic Compound arid Hazardous
Air Pollutant Emissions. To calculate model unit baseline and
MACT VOC emissions, the appropriate emission factors are
multiplied by their respective equipment counts. The VOC
baseline emissions, MACT emissions, and emission reductions
are presented in Tables 6-5 through 6-7, respectively.
Corresponding HAP emissions for each model unit are reported
at the bottom of each table, and were calculated using the
0.78 average HAP-to-VOC ratio. Model baseline emissions of
VOC's range from 26 Mg/yr to 355 Mg/yr. Model MACT emissions
of VOC's ranged from 5 to 48 Mg/yr.
6.3.2 Secondary Environmental Impacts
6.3.2.1 Water Quality. Reduction of VOC cind HAP
emissions from equipment in liquid service may result in
reduced loading to wastewater streams. However, the nature of
these materials is that they evaporate to the air. Overall,
the impacts, both positive and negative on wastewater, from
the negotiated regulation5 would be minor.
6.3.2.2 Solid Waste. Solid waste from SOCMI pertaining
to equipment leaks includes replaced seals, packing, rupture
disks, and used equipment components such as pumps and valves
that have been replaced. Metal solid wastes such as
mechanical seals, rupture disks, and valve parts could be sold
to companies that can recycle the metal.
6-18
-------
CO
H
H^j
"~^
pa
o TT
g £
/^ bO
r-H ty>
O C^
fa
M
CO C
Sz °
o '£
CO NH
CO E
r-l ^
M
W O
>
rf! 0)
hr-J £•«
MH «r-1
r-4
Q M u
3 0
E— '
U
u -a
•U v. r-4
•rj g r-l
£ rH °
Oi
-H g C
CU ,
C C
Q) -, 01
w 43 F> 0) w
hO cd 5-1 CU
3 o cd
CM O >
o o o
VO CO VO
r-l CO in
. *
r-l CO
r-l
o o o
in in vo
VO rH in
r^ co CM
VO CM
1-1
o o o
r-4 CO j.
in
CO
CO
^.'
00
r-l
CO
VO
OO
VO
O
CM
m
VO
CM
VO
o
co
r-l
-*
CM
CO
P Emissions
g/yr)
3~
*
CO
H
0)
O
0)
r-l
0)
(U
CM
£
•H
Tf
0)
•H
ft I
•H
O
0)
ft
CQ
rH
0)
rH
d)
^
O
-P
Tf
0)
,_(
H
O
H
-P
O
O
II
T3
0)
rH
_ 1
TG-contro:
u
Q)
C
•H
•H
CD
U
rri
lu •
Q Q)
C
-P -H
(0 rH
0)
*O 05
o) (d
rH X)
rH
O 4J
rH (0
-P
d -d
O 0)
O H
H
o o
B rH
U 4J
Q) O
XI 0
O O
•P £-<
U
-0
CU el)
3 Xt
^j .
w o
W 4->
rt!
Q> 0)
rH S
(0 3
CQ
CQ CQ
o) (d
rH 0)
It! rH
^ Tf
eventy-fi1
ne hundre(
CO O
XI 0
-------
TABLE 6-6. MODEL UNIT VOC EMISSIONS FROM MACT CONTROL
Equipment
Component
Pump seals
Light-liquid
Heavy-liquid
Compressor
seals
Valves
Gas
Light-liquid
Heavy- liquid
Pressure
MACT Control VOC Emissions (kg/yr)
Model Units
A and D
Simple
330
0
0
80
2,390
0
0
Model Units
B and E
Medium
870
940
390
440
7,410
140
0
Model Units
C and F
Complex
1,220
6,750
1,560
1,470
12,440
2,560
0
relief devices
Open-ended 0
lines
Sampling 0
connections
Connectors 2,210
0
0
8,050
0
0
22,390
TOTAL 5,010 18,240
EMISSIONS
(kg/yr)
TOTAL 5.0 18.2
EMISSIONS
(VOC)
(Mg/yr)
HAP EMISSIONS 3.9 14.2
(Mg/yr)
48,390
48.4
37.7
6-20
-------
en
o
H
EH '
U
D
Q
P£
EMISSION
ft
H
B
t_^
H
Q
§
•
1
VO
W
CQ
^
—
-
^^
j_j
>-,
"fan
missions (kj
w
o
o
CU
c
•r-4
rH
CU
01
cd
CO
fl
fa "°
» Q)
[
'g CO O
| — I rj^ jj
B C
* O O
"o rj,
S H
cd
Cd "°
W CO
III
3*1
i 8
0
cd
Q "5
Model Unit
Simple
TG-Controll(
o
O T3
CU
£4 i O
"OB ° §
•o u o
0 C
s &
CU
4-1 rH
•H S rH
C J 0
T) 4J
rH CO f3
0) S 0
•O 0
0 C
2 D
^^ 13
CU
•H <0 rH
C rH 0
CD co o
TJ O
O C
2 P
1 i 1 1
C C
CO CU
S C
a. o
II
- - -• -wo
o o
CM
o
m
i — i
0 0
CO
0
rH
O O
CM
0
W >,
W rfl P
t>o cd
ex, .H -J x
o
VO
rH
CTi
O
C^
CM
CM
O
0
CM
o
CTv
VJ3-_
CM
CM
0
rH
OO
VO
O
r-^
CM
••
rH
O
OO
•rH
VO
VO
O
00
C^
rH
O
0
r**.
CO
01
cd
a
0
N^
"I
VO
m
0
0
cr>
co"
CO
o
co
CJ\
~
o
o
rH
1^
0
i-H
rH
O
CM
0\
in
VO
o
CM
rH
CM
T3
•rl
0*
-i-4
rH
M
•rH
_1
O
0
o
o
o
o
•rj
•rl
cr
•H
rH
cd
CO
33
0 0
m
f"-
oo
CO
0 0
H *O
CO
0) T3
>H 01 C
3 CO CU
CO O 1
01 -H C
CO > CO
JH cu CX
PM "O O
o
CO
_os
CO
rH
o
o
VO
^jT
o
in
o
-
rH
0
CO
O^i
CO
rH
O
0
VO
^jT
o
m
o
rH
oo
•H
rH
CO
CO
O
•H
4J
0
conne
0
O oo
CTi CO
0- rH
rH
CO
VO
O CM
o o**
CO
rH
rH
O rH
O CM
rH
K
co
rH
O VO
CTi O
<)• CO
rH
CO
CM
O O"v
0 •
M « —
CO
0] CO CO
rJ M Z
0 2O
4-) W M x-x
0 H SH
S g|^
VO
VO
rH
CM
CM
VO
rH
00
oo
CO
CM
^
v^
oo
o
M CO
CO K
CO O
M I-H
S-* r i
P-l Q
^ w
CO
r-l
Q)
O
(U
M
0)
-------
Although additional monitoring of equipment may result in
a greater rate of replacement for faulty equipment, it may
also reduce equipment failure. Overall, no significant impact
on solid waste as a result of implementing the negotiated
regulation is expected.
6.3.2.3 Energy. The measures taken to comply with the
negotiated regulation5 will result in reduced emissions of
organic feedstocks, intermediates, and products. Increasing
the efficiency of production at SOCMI facilities will result
in reduced use of chemical feedstocks. Additionally, this
higher efficiency will result in energy savings while
producing materials that otherwise would be lost.
6.4 COST IMPACTS OF CONTROLLING EMISSIONS FROM EQUIPMENT
LEAKS
The following sections discuss the methodology used to
develop capital and annual costs for the control of HAP
emissions from equipment leaks. These costs are based on the
base costs per component, which were presented in Section 3.4
of Volume IB, and model unit equipment counts, which were
presented in Section 6.2 of this volume.
6.4.1 Capital Costs
The discussion of capital costs for the control of
emissions from equipment leaks has been divided into two
sections. The first section covers all capital costs for the
purchase and installation of control equipment for
compressors, pressure relief devices, and sample connections.
The second section covers all capital costs for the
implementation of a monthly LDAR program for pumps, valves,
and connectors. This includes the purchase of one portable
hydrocarbon monitoring instrument (organic vapor analyzer) and
the costs of an initial individual component survey.
6.4.1.1 Control Equipment. The control equipment
required for compressors, pressure relief devices, open-ended
lines, and sample connections were discussed in Section 2.4 of
Volume IB. The base costs for equipment for these types of
equipment were developed and presented in Section 3.4 of
6-22
-------
Volume IB. The total installed base costs per component for
each type of equipment are shown in Table 6-8. Using these
base costs, and the appropriate equipment counts for each of .
the six model units, the capital costs for control equipment
in each model unit are presented in Table 6-9. There are no
costs associated with the installation of caps for open-ended
lines in any of the model units. All model units are assumed
to already have caps installed on all open-ended lines.
6.4.1.2 Initial Leak Detection and Repair. The capital
costs associated with the initial LDAR program are the initial
purchase cost of a portable hydrocarbon monitoring instrument,
which is included in Table 6-9, and the cost of the initial
individual component survey for each process unit.
Those process units that already have an operating LDAR
program will not incur these capital expenses. Each model
unit that does not already have such a routine program will
incur these capital expenses. Model units D, E, and F, are
assumed to be controlled to the levels recommended in the
SOCMI equipment leaks CTG,17 with an operating LDAR program.
Model units A, B, and C, are assumed to be uncontrolled, and
will have the added capital expenses of initiating the LDAR
program. The initial individual component survey costs for
model units A, B, and C are developed in Table 6-10. The
sources of information and assumptions made to develop this
table are discussed in detail in Section 3.4 of Volume IB.
The total labor costs for the initial individual
component survey are $6,640 for model unit A; $22,838 for
model unit B; and $53,578 for model unit C. These costs do
not include the cost of replacement pump seals. The cost for
replacement pump seals is $180 per seal, or $405 for model
unit A; $1,080 for model unit B; and $1,512 for model unit C.
Replacement seals for valves and connectors are considered to
be covered by routine plant maintenance and are not included
in this analysis.
The total capital costs for the three model units that
require an initial survey, including costs for initial
6-23
-------
TABLE 6-8. TOTAL INSTALLED BASE COSTS
Equipment Type
Base Cost
($)
Compressors
Pressure relief devices
• Rupture disks
• Holders, valves, etc.
Open-ended lines
Sample connections
6,500
78
3,852
102
409
6-24
-------
co
EH
H
fe
O
55
H
EH
W
H
X
W
O
fa
EH
S
H
g
H
B
W
tf
0
fa
CO
EH
CO
O
U
a
EH
H
PH
1^
U
*
^J^
vo
H
CQ
(d
f*
o
-H
-P
fd
o
-H
((H
•H
-P
C
0)
73
H
-P
•H
C
§
rH
CD
'O
*
fa
Q
U
CO
4
«•—•».
4fa
N— '
-P
01
O
u
45 O rj\ n 45 CM n
O •*}• O in O rH H
CJ> rj\ P* CO fj>
CT> in oo • co rH
CM CO
45 in oo o 45 H o
H CM
45 U in t 45 a\ oo
o o in o o vo CM
H r> CM rH
•». ^ *.
c~» con
rH
O O CO 45 CM CO
O •* O tn O tH H
vo CM VO
vo r- co p*
O 01
• MJ f*
rH C
•P O
^ *l
0) HO
-P rd 01 o>
01 4J Q) C
>> in 01 c c
01 Q) C -H O
4J -H -H H O
C 4J H
G G
-P H 01 H 0) rd
01 fd fd fd I 01
CO) > C
•H 01 r* 0) 0)
01 - ft CT>
C7> rH -H 01 O rH
C 0 T3 rH 3
•H 01 01 Q) }H ft
rH 01 0) ^! -C O 1
O 0) rH 01 H <»H 73
-P rH S -H O Q) 1-5
•Hft4JQ330101i<
C g ft ft 0 EH
o o 3 i i id H o
g U « 1 1 U U EH
W -P
-P C
•H Q)
CO H >
3 -P Q)
T3 H
H -p O fd
0) C g . Q)
T3 Q) r~' 01
OH HO
g Id H rH M
> < -P O
•H C (/]
: ^fc w*
• J3 O rjl
WO1 • O Q)
-P 0) 01 rH
•H +J^ ft
C 01 CO' g
3 H (D EH 01 O
O g O g O
T3 )H 30)
0) +J rH -P -P rH
H C -P Q) -H O
HO 01 0) H
ft-H T3 -P O
g £ Q) C (d <4H
O > i ft 01 -P
fd 3 73 O rd
-X 0) C 0) Ol"
gO) rH O rH rH
3 H H (d O
•H ft (d 01 01
73 g ft 01 01
Q) O O fd -P 0
gO -P U W rH
O ft
•«. "73 T3 T3 O €
(DC 0) 0) O
H fd g H H O
ft 3 H rd
& *. 01 fd C cu
-H g 01 -P O >
01 3 fd 01 -H rd
•H C -P 43
0) T3 0) -H -H
rH 0) rH 73 -P
fd g td a) 73 o
> fd C
o •* fo td
Q)S 43 O O
T3 <~H • 13 C- 'O
c ft 01 co
(d g H fd 4J ~ Q
•HO) Q)
•> 01 > «-73 rH 73
CQ 0) H CU O C
0) H g
73 O 73 "H J> 73 01
O-43 OCO) O>i
gQ4J g3H gOl
fd 45 O
6-25
-------
I
H
55
O
a
H
H
I
vo
03
CO
o
(J
1
on
CD
V
c_
CD t-
c- a
.— CU
CD o:
CU CO
°e 3
^y O
S =
Of
o
m
<
cj
CO
•~° "M
v c <-o
C— 0) C.
4) «- .c ••-
O <•- 4-> CO
t- 3 t- Q.
o cr 3 d>
Q. 5 u. oe
0£
M-
o
L.
5 CO
•s!
4-»
ffl
.1
CO
LU
u
CO
*
' TS ^
T« "CD ^
•g «} p-
•"• u
u.
CO
CO
8
u
en
C
t.
o
|
o
m
c
•si
a* °
x
1
CU
u
I
CO
f»- *o
to o
in co
to CM
CNJ ^a-
-
f*-. co
in CM
*~ *~
IM >O
N- *O
•* N-
co r-
co -j
CM
..*
in in
CM CM
CM N-
r^ co
in CM
(M -O
-* N-
CO N-
"'a
-a- in
J- 0
o a
CO O
*a- m
to"-*'
in co
sa
"~ ^
Nl O
O in
4-1 4-*
C C
5 5
U L.
4-* 4->
CO CO
C C
TJ
<-!
CD
> 4J
co •>, j:
OJ CO O>
> CO ••-
— • 13 _1
>
a
0
to"
I
CM"
o
CO
2
•~
o
•o
o
o
•0
to
s
in
t-.
CM
JI
O
co
o
M
CM
O
O
«~
O
0
CO
to
4J
5
u
4J
8
•o
'1
CO ••-
(Q
CO -C
0)
o.
s
^~
0
s
£
CO
fc
0
CO
CM
h-
1^
OJ
in
OJ
NO
in
in
in
in
^
in
r\l
o
d
tn
CM
in
CO*
^"
in
in
NO
NO
CO
c
5
t_
(0
c
CO
0
u
I
o
u
§
CM
«~
NO
IN-
tn
tn
R
to
o
rs.
CM
CO
R
o
*~
CO
o
o
CO
^
1—
o
m
-
U OJ
•5°-
O > «- "-
O> d}
3 TJ
c.5
CU I—
CU
a> i
CD i
CU
•3
•O
O
*4-
0
8 '5.
CO
2 .1
•a w
CD
TJ
D> CU
C N
Z —•
O CD
+•* U
c E
O '^
Q.
3t tO
- "8
gi
CU CU
a. t-
u
co
8
•g
CD
g
4-*
ta
L_
4-*
CO
'E
1
1
u
CU
t_
CO
CD
4-*
O
6-26
-------
individual component surveys and replacement pump seals, are
$24,270 for model unit A; $232,970 for model unit B; and
$451,990 for model unit C. The total capital costs for model
units D, E, and F, are $11,130; $203,630; and $391,910,
respectively. These are the same costs presented in
Table 6-9, since these three model units already have an
operating LDAR program.
6.4.2 Annual Costs
The discussion of annual costs for the control of HAP
emissions from the six model units has been divided into five
sections: annualized capital costs, maintenance charges,
miscellaneous charges, LDAR costs, and recovery credits. The
method for the development of annual costs is discussed in
detail in Section 3.4 of Volume IB. Each of these sections is
discussed briefly below.
6.4.2.1 Annualized Capital Costs. The annualized
capital costs were calculated by taking the appropriate
capital recovery factor as discussed in Section 3.4 of
Volume IB and applying it to the corresponding capital cost
from Table 6-9.
6.4.2.2 Annual Maintenance Charges. The annual
maintenance charge for control eguipment was calculated by
multiplying the appropriate capital cost from Table 6-9 by
0.05. The annual maintenance charge for the portable
hydrocarbon monitoring instrument was $4,280 per year. This
cost was updated from the Fugitive Emissions AID.12
The costs of replacement pump seals under the LDAR
program were considered to be a maintenance expense. They
were calculated by multiplying the replacement seal cost of
$180 per seal by the number of pump leaks repaired annually.
6.4.2.3 Annual Miscellaneous Charges. The miscellaneous
charges for control equipment and for the portable VOC
detection instrument were calculated by applying the factor of
0.04 to the appropriate capital cost. The miscellaneous
charge for replacement pump seals was calculated as 80 percent
of the annual maintenance charge for pump seals.
6-27
-------
6.4.2.4 Annual Leak Detection and Repair Costs. The
annual operating costs for the LDAR program will vary
depending on the number of equipment components to be
surveyed, the monitoring frequency, the leak frequency, and
the cost of leak repairs. The annual monitoring and leak
repair costs, including administration and support costs, used
in this analysis are presented in Table 6-11. Administration
and support costs for a LDAR program are equal to 40 percent
of the monitoring and leak repair costs. Costs in Table 6-11
are based on a monthly valve monitoring frequency. If a
process unit maintains valve leak frequency below two percent,
valve monitoring will have to be performed less frequently,
and the monitoring and repair cost for valves will be
proportionately less. This is important since valve
monitoring and repair cost are a significant portion of total
annual cost. If valve monitoring frequency were quarterly
versus monthly, total annual monitoring and leak repair costs
would be reduced by 45 to 50 percent depending on the model
unit.
6.4.2.5 Recovery Credits. The recovery credit values
for each of the six model units were determined by multiplying
the annual VOC emission reduction, as presented in Table 6-7
by the average chemical cost of $l,590/Mg of VOC to produce
the recovered raw material credits. The average chemical cost
was determined based on available cost data for 168 VOC
chemicals.
6.4.2.6 Total Annual Costs. Total annual cost is the
sum of annualized capital cost, maintenance charges,
miscellaneous charges, LDAR costs, and recovery credits where
recovery credits have negative values representing cost
savings. Table 6-12 summarizes these costs for the six model
units.
6.4.3 Cost Effectiveness
The cost effectiveness for each of the six model units is
presented in Table 6-13. The values in Table 6-13 reflect
monthly valve monitoring.
6-28
-------
g
O
EH
H
Si1
i
H
I
vo
I
4-*
CO
o
§"~
Of
\S
to
CU
_J
1
OTJ
m ?
_j.;_
.^ Sf
(D OC
Q.
CU CO
3
•££
cu
u.
tj
UJ
m
a
<
u.
<->
UJ
m
Q
^
Q. ~ ^
*> £ 5
at1—^
CO
C -H 01 t-
QJ L. _C >»—
O "- 4-* ID
£- 3 1_ O.
.
c
S3
P"
— 1 CU
c_
u.
CO
yj
CO
u
OI
c
t-
o
4-»
1
u.
•v.
U
UJ
m
1
c
•*- .*-
o £
H
i
01
u
i
CO
f- CM
>T t7>
•»t y
^
-O y
CM r*
CM -O
^O Oi
*— in
si- C
o.
o f\
SIC
ro **
•* o
c>
in c\
co ^
„
in in
CM CM
O CM
*~ in
ro N.
«* o
~
in CM
CM CM
o o
o o
-o o
O CM
o in
10 ^>
ro •»
ro 0?
CM
35
4-* 4-»
C C
^ K
t_ c_
4-« 4-*
CO CO
C C
c-I
8.=
M§.M
OI CO O)
> CO —
5
"~.
2
o
§
CM
.- o
s
^~
-*
e?
OO
0
S
in
o
<0
IM
>0
1C
CM
JS^
»O
O
CO
•J-
0
CO
«—
o
fi in
v—
§0
o
ro
0 -0
£
02
c- "co
4J 3
CO CO
'3
«T
CO "-
CO
01 4->
o
co
^
in
CM
,-
|S^
ro
ro
ro
in
CD
O
O
o
CM
CO
-*"
CM
in
S!
%
O
4-»
1
o
Connec
0
o
o
o
o
o
0
0
o
o
o
-
o
0
o
o
o
o
o
CM
in
o
o
-
4-»
c
Q
£_
4->
1
*4-
0)
0
313
CO M
CO — •
01 CO
C- 01
a. co
1
£
CM
^
CM
o
•<
[S.
s
rC
o
c>
>»
o
s
*~
tn
*£
SUBTOT,
N-
«-
S
^~
CO
e
y
-o
§
O
OJ
CO
4-»
CO
o
CJ
C-
CO
li
QJ
0£
•g
CO
CO
c
L.
O
'E
o
4-1
o
JT-
"
I/
ro
a
c\
o
a
cu
CO
M
cS
L.
|
CO
•52
c
CO
c
o
L.
CO
'E
i
00
a
a
*~
,_
t*
.8"
ro
K
K
c\
-
t/t
t—
C/}
8
1
1
o
1
8
CO
CD
ro
c
o
(U
CO
CO
01
c_
CD
CO
4-f
S
o
g
o
._
CD
3
CO
c
OI
01
c
c.
0
4-»
'1
4-1
1
0
0 ll
0 3
• O
CM -C
*• S
o a.
01 o
01 in
M- •
0 **
4-* «f-
0 O
il
.2 co
CO 4->
CO
CO
4-*
C C
O 01
"S "
CO CO
t-
01 0
C_ 4-*
CD —
c
CO Q
I-
U CD
t» CO
C —
O t,
£8.
c
if
I- U
I fc
CO
i
o
4-*
c
01
u
8.
1C
•£
M-
o
c
o
'Z •
OI >.
U 0)
X >
01 C.
OI CO
JZ
** c
4-1 — '~
— Q.
X
CM C
o -a
»~ CD
< 4->
CO J3
> "U
•^ £_
4-> CO
1-8
X
i_ o
g 0,
o c
en 01
01 X
JZ J=
4J 4-»
01
l§
H- C_
c
OI CO
^ V
CO CD
4J T3
2 E?
§•§
CO
113
• .- 4-<
4-> CD
t. 3
s -
a. co
01 U
•gi
CO
CO CD
a. >
OI
III
OI 1—
J=
£. .
3 c_
C ll
'i-l
01 4J
U f.
CO <4-
01
CO CO
CO C
c 'H
01 —
o cu
J3
L.
-C
o ,
in
CO
OJ
iA
CD
CO
CO
o
o
1
CO
*-
8.
OI
c_
01
_3
g
CO
4-*
CO
O
u
f_
'5
a
0)
1
_J
U
t-
01
OI U
•o c
CD
S^
g;
f U
O c-
-C 3
4-* CO
•o
,
CO
CO
0
u
L.
5
s-
L.
CD
O
c
L.
O
4-*
'E
g
CD
O
4-*
«4-
0
S
o
c_
8.
O
"S
4-*
CO
O
U
4^
C_
g-
w
•g
CD
C
O
CO
CO
'E
i
«c
Cl*
6-29
-------
u.
JJ
C
53
H
O
EH
g
O £
W
1
| :
EH c
55 =
g
CO
EH
5
^ .ti
c
•4 =
W
a
o
£
g -
CO 'c
EH =
co
O
U
0
EH <
S ~
O 'c
0 =>
1
53
*5
i^n
oa
H
1
vo
W
CQ
E
c
4-1
4-*
(A
O
O
«— NJ- N- o ro
co" roW N^" *Q
**$• *o
ro CM in to in
o o CM^ ro^ "O^
CM CNJ eg CNJ p*
o o o *o o ro o o c>
o in ro t-
CM in co
«-T «-?
m o Is* O** o o ro [^ ^~
J »M O*
^— CM fM CO CM Kl ^
CM -*
*
mo o^ooro M SG 5T
o o in ro co fo o»
•Cj CM in CD CM ^o
»- T- t- ^*
g
4-1 W 05 •*-*
O) +-1 OJ £-
>*(/)***£ O
g4J+j!^w~"'o> 3?
c o 5? ^? ^ oj o>
§fl) gS > QJ i^- O)
> S — ' TJ — • "3
*- O) (0 (U ' tO *•" 3
CO 4_» C flJ * C U) (D • U.
j=c— «^r3!«P-c
CJoj (fl'-O.Q.flJ ••—•*-'
_^ g^ o L. •— ^ 0§ *" -S i
(0 i^ .*— y) •"* ^ ^ .^ (0 E
4-*-3t-(Qfl)Oo>tO O tfl
*S IT O 01 t. ,- -C st. -o fl» ~" S
UQi4_ft_3w OJ — »t-ra TT
ra.,-Q.ii-^-^u)M o — • S
u _, c g o.." .^ Q- ° — • -g g- cj
"g SzcjaOoSo-SSo- ..
N c ' ' •— "3 .£
"Sra^jo -o «; — 10 ^5 •g '^
E ^ o
3 Q
«£ ,_ CM CO O
CM ** CM s *c
o o co u% co ro o
ru r-i «- *- -st
*jr" CM
OS- in 03 in NO CM
co c> o ro in ^ S-
CM stf- CM ^D «— ^- O
•^r" CM ^i- CM CM O*
o xf in t1- S- in o
CO CM N- \0 v- J^ CC
-^" CO <
§0 co in co ro o
co o S3 ro
PJ pn «— «- ^3-
^ c\
e
•M U CO
(A ,(-» flj
c/i to
-------
o
55
H
O
M
^
O
^
S3
EH
£5
O
EH
H
1 MODEL
:LUDED;
go
m o
EH
0
U
O
EH .
O
O
O *O CO CO rO
§in*
V
CM CM
ooj-^rocNJ^r ^ot\iro«— ^>tn
(M i* O in IA «— O*h*'»O(M«— *O
h- in -^«— ^oo-h-^o
•>* e\j CNJ c\i N* «—
O O ^^ -J" «-• "^ O (^ ^f (3 O CK
rvi ro ^~ ^~ o^ ex ro ^~ ^3 o in
*— c\j r>» co co ">^ >o
«- ro ro
OCO'O-^CMCO co t> «— o in •»»•
^OO^tPJfOin N-ON-N-^-f>
CM 0^ CK r>*» 1s" rvj *>^ '•o "O oo r>» ro
*—«—«—«— N- N->OfON-*O«—
«- o ro in ** co >o
CM -^roc\j%a- NOtMro^rur^
^O O^ r*>» P1* p» OO Q* T— CO «"• T~ ^D
e\ivj-oinin«— o^tv-NCicMCMin
r>» in ^^ ^~ *o ex r>- f>-
<^ c\j ru CM ro *o
«— C\J v^
CO
^o «~roro-^ •>*^CNJincO's±
CNJ ro^-«— o oro«-^ocx^o
^~ CNI P«" OO CO f>- *«^
«- - ro in «—
o •*-•
•M CO C.
- CO 0> O
CO >» C Q-
-g _* c- c
« c_ u CQ.O. -r-co
ococora i to 5 CQ.OI
cu><-c
CC"-tfl-^tO t_L_ —
CO ••- (A Q. Q. -M O *-•
= -§g&:508&»£SS
aico •*-(/! (- — ' £ cu c QJ 4-* .
u i-cnoo.ScnoDw "
M-o4^i-3 cumcD cue '^=
EO'^Q.4-»eflc/s— »JIQC «)•»- *
CCe^Q.QLOQ.CJ<^£ ^*
— 'CDOO3CO — 'O QI3"O ,4
§c_zcjQscjejQ:t-_ i«<-1tj
w "CD L
<— C0^3 OTJ 01 H- _1 CO^J "S >_l
4-> O <
-Q 0 I—
3 01 O
CM ro cn cc i—
i
c
CD
U)
a>
ca
S
CD
01
.
CD
Ol
0)
CD
U
in
en
o>
.£?
4-*
C
CD"
6-31
-------
£T
H
o
EH
H
§
a
w
!j
>
>H
1
O
S3
CO
EH
rH
53
D
f~4
w
Q
O
SB
PH
o
EH
CO
CO
w
w
H
s
w
P-4
Pn
W
EH
CO
O
U
•
CO
H
1
VO
W
3
EH
r-g.
-P
•H
&
..
•P
•H
c*1
5
Q
•P
•H
C
£3
u
-p
•H
|-^
fr«J
CQ
•P
•H
C
^
1 1
•H
•^
t*.
CM
CO
in
,_!
in
VO
vo
H
en
rH
in
*•
^^
in
en
rH
.
^,
O
in
%.
vo
a
VO
^4
vo
H
*^s
•ci
0)
N
•H
cd -^^*
C >i
ctJ
JJ W
0 O
EH 0
co
en
•
F^
CO
H
r^
in
en
en
CO
rH
CM
,_)
H
VO
O
CO
en
rH
•
en
•<*
rH
[N»
^
VO
CO
10
C
O
•H
-P
O
rQ
Q) ^
O O
-H O
w >
co
-H &>
W --'
CO
t~»
•
0
CO
c^
o
CO
[^
o
CO
.
o
CO
f^
.
0
CO
o
u
o
•»,.
P-4
Q)
nj O
JH-H
Q) -P
> cd
CM
VO
*
VO
••*
H
O
CM
CM
CO
VO
VO
rH
P^
.
co
CO
CM
f^
CO
VO
rH
rH
in
"^
CO
CM
CO
C
O
•H
•P
O
f3
13
0) —
li t t
c ^~.
O PM
•H ril
(0 E
01
•H &>
IS
0
co
o
CO
rH
0
0
CO
,— ,
o
CO
in
^_^
o
in
^f
^~*
^
o
o
^J<
^^
to
10
0)
c
0)
^
•H
-P
O ^
Q) CJ
OH O
OH >
0)
-P S*
CO --.
o
CJ — '
0
o
1-1
o
CO
CM
o
en
CO
,— ,
o
CO
VO
^^ ^
o
co
in
_
o
rH
in
*--'
CO
CO
0)
c
0)
^
•H
-P
O ~
(U Oi
OH ffi
0)
-P S
co --.
o
u —
*
"to"
CP
c
•H
^
(0
w
^
•
Q)
-H
**•• •*
CO
0)
^
rH
cd
>
Q)
^
•H
-p
&
(1)
c
QJ
-p
ra
o
•H
T3
C
•rH
CO
0)
CO
0)
XJ
4J
CU
JH
(0
PH
rd
6-32
-------
Cost effectiveness was determined on a VOC basis and a
HAP basis. The total annual costs were divided by the VOC and
HAP emission reductions giving the VOC and HAP cost
effectiveness, respectively, for each model unit.
These calculations indicate a net savings for model units
A, B, and C, and a net expense for model units D, E, and F,
with monthly valve monitoring. For the monthly valve
monitoring frequency, the cost effectiveness values range from
a savings of $650/Mg HAP for model unit C, to an expense of
$390/Mg HAP for model unit D.
6-33
-------
6.5 REFERENCES
1. Erikson, D. G. and V. Kalcevic (IT Enviroscience, Inc.).
Fugitive Emissions. In: Organic Chemical Manufacturing,
Volume 3: Storage, Fugitive, and Secondary Sources.
Report 2. U. S. Environmental Protection Agency.
Research Triangle Park, NC. EPA-450/3-80-025. December
1980. p. II-2.
2. Ref. 1, p. II-6.
3. McFarland, I. Preventing Flange Fires. Chem. Eng. Prog.
65(8):59-61. August 1969.
4. Ramsey, W. D. and G. C. Zoller. How the Design of
Shafts, Seals and Impellers Affects Agitator Performance.
Chem. Eng. 83(18):101-108. August 30, 1976.
5. National Emission Standards for Hazardous Air Pollutants;
Announcement of Negotiated Regulation for Equipment
Leaks. Federal Register. Vol. 56, No. 44. Washington,
DC. Office of the Federal Register. March 6, 1991.
6. U.S. Environmental Protection Agency. Fugitive Emission
Sources of Organic Compounds—Additional Information on
Emissions, Emission Reduction, and Costs. Section 3.
EPA-450/3-82-010. Research Triangle Park, NC. April
1982.
7. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Benzene Fugitive
Emissions—Background Information for Proposed Standards.
Research Triangle Park, NC. Publication No. Chapter 6.
EPA-450/3-80-032a. November 1980.
8. Blacksmith, J. R., G. E. Harris, and G. L. Langley
(Radian Corporation). Problem-Oriented Report:
Frequency of Leak Occurrence for Fittings in Synthetic
Organic Chemical Plant Process Plants. Prepared for
U. S. Environmental Protection Agency. Research Triangle
Park, NC. Publication No. EPA-600/2-81-003. September
1980.
9. Memorandum from Moretti, E. C., Radian Corporation, to
Markwordt D., EPA/CPB. May 22, 1989. Equipment leak
emission estimates, VOC/HAP.
10. Ref. 6, p. 2-70.
11. Ref. 6, Section 4.
12. Ref. 6, Section 5.
6-34
-------
13. U.S. Environmental Protection Agency. Emission Factors
and Frequency of Leak Occurrence for Fittings in Refinery
Process Units. EPA-600/2-79-044. Research Triangle
Park, NC. February 1979.
14. Langely, G.M., S.M. Dennis, L.P. Provost, and J.F. Ward
(Radian Corporation). Analysis of SOCMI VOC Fugitive
Emissions Data. Prepared for U.S. Environmental
Protection Agency. Research Triangle Park, NC.
Publication No. EPA-600/2-81-111. June 1981.
15. Pullman Kellogg Company. Equipment Component Analysis
for Identification of Potential Fugitive Emission
Sources. Prepared for U. S. Environmental Protection
Agency. Research Triangle Park, NC. June 1978.
16. Ref. 1, p. 11-13.
17. U.S. Environmental Protection Agency. Control of
Volatile Organic Compounds Leaks from synthetic Organic
Chemical and Polymer Manufacturing Equipment. Guideline
Series. EPA-450/3-83-006. Research Triangle Park, NC.
March 1984.
18. Ref. 17, p. 3-8 and 3-9.
19. Memorandum from Hausle, K.J, Radian Corporation, to
Markwordt, D., EPA/CPB. February 21, 1992. Emission and
control impacts estimation methodology for small process
units for HON—equipment leaks.
20. Memorandum from Hohenstein, W.G., Radian Corporation to
Markwordt, D., EPA/CPB. March 29, 1991. Revised "reg
neg" controlled model unit emissions from equipment
leaks.
6-35
-------
-------
APPENDIX A
EXAMPLE IMPACTS FOR APPLICATION OF FLARE ON A PROCESS VENT
A.I INTRODUCTION
This appendix presents example calculations of primary
air pollution impacts and cost impacts for controlling process
vents by combustion with flares or incinerators. The process
vent discussed here is identical to model process vent 7
presented in Tables 2-6 through 2-9 of the text.
Because vent streams vary in flow rate and composition,
it is important to note that the emissions and cost impacts
from the example process, vent stream do not represent the
impacts for all process vents in the SOCMI. This process vent
was selected only to illustrate the calculations necessary to
estimate the impacts of controlling process vent emissions in
the SOCMI.
The purpose of this appendix is to demonstrate the
approach used in the HON analysis. In the calculations below,
all significant figures have been retained until the final
calculation to make it easier for the reader to follow the
calculation and to avoid potential error due to round off of
intermediate calculations. It should not be inferred that the
intermediate results represent the actual number of
significant figures.
A.2 MODEL ASSIGNMENT
Models representing vent stream characteristics were
assigned to facilities based on the chemical production
process. The example stream is a process vent from the
production of ethylbenzene via alkylation of benzene with
ethylene. Specific stream data were available for this kind
A-l
-------
of process so it was possible to assign a Type A model to this
stream. Further details on model development are presented in
Section 2.2.2 of the text. Although Type A models were
assigned to both the reactor and distillation vents associated
with this process, this appendix only addresses the
distillation vent, which was assigned the ethylbenzene model
stream from Table 2-3 in the text. Table A-l presents the
ethylbenzene model stream parameters, and Table A-2 presents a
summary of calculated results for the example stream.
The stream data generated from the model was based on a
reported facility production capacity of 405 Gg/yr. Values
were generated for flow rate, VOC emissions and HAP emissions
based on a ratio of the process capacity to the model capacity
as follows:
Example Value =
Flow Rate
Uncontrolled
VOC Emissions
Uncontrolled
HAP Emissions
Model Value * (Production Capacity -.
Model Production Capacity);
(2.9 scfm) [405 Gg/yr •*• 159 Gg/yr] =
7.38 scfm;
(1.33 Mg/yr)[405 Gg/yr
3.38 Mg/yr; and
159 Gg/yr] =
(1.33 Mg/yr) [405 Gg/yr * 159 Gg/yr] =
3.38 Mg/yr.
Because heat content, temperature, oxygen content, and VOC and
HAP composition are intrinsic properties, they are not
affected by the production capacity of a process and would not
be scaled up like flow rate and emission rates.
A.3 BASELINE EMISSIONS
The calculation of baseline emissions was used to
estimate the actual emissions after a control device required
by a county, State, or national regulation was applied.
As shown in Table A-2, the example stream is located in
Galveston County, TX. The applicable regulation for this
county requires a control device if VOC emissions are greater
than 100 Ibs/day. Assuming the vent operates 365 days/yr, the
example vent uncontrolled emissions would be 20.4 Ibs/day, so
A-2
-------
TABLE A-l. MODEL STREAM PARAMETERS
Parameter
Unit
Distillation
Number of Columns
Distillation Type
Halogen
Capacity
Flow Rate
Heat Content
Temperature
C>2 Content
VOC Emissions
Vbc Composition
HAP Emissions
HAP Composition
Gg/yr
scfm
Btu/scf
OF
volume %
Mg/yr
weight %
Mg/yr
weight %
2
NV/V
N
159
2.9
321
110.7
0
1.33
2.7
1.33
2.7
A-3
-------
TABLE A-2. EXAMPLE STREAM PARAMETERS
Parameter
State
County
City
Flow Rate
Uncontrolled VOC Emissions
Uncontrolled HAP Emissions
Baseline VOC Emissions
Baseline HAP Emissions
Controlled VOC Emissions
Controlled HAP Emissions
VOC Emission Reduction
HAP Emission Reduction
Total Annual Control Cost
Cost Effectiveness
Unit
—
—
—
scfm
Mg/yr
Mg/yr
Mg/yr
Mg/yr
Mg/yr
Mg/yr
Mg/yr
Mg/yr
$/yr
$/Mg HAP
Reduction
Distillation
TX
Galveston
Texas City
7.38
3.38
3.38
3.38
3.38
0.0676
0.0676
3.31
3.31
46,115
13,932
A-4
-------
no control device would be required, and the baseline
emissions (3.38 Mg HAP/yr, 3.38 Mg VOC/yr) would be equivalent
to the uncontrolled emissions calculated in Section A-2 of
this appendix.
A.4 CONTROL DEVICE ASSIGNMENT
Three control technologies were evaluated for process
vents in the HON analysis - flares, thermal incinerators, and
thermal incinerators with scrubbers. A thermal incinerator
with a scrubber was only considered for halogenated streams.
When streams were not halogenated, either a flare or thermal
incinerator was chosen on the basis of cost effectiveness
($/Mg HAP removed). For each nonhalogenated vent stream, a
flare and four incinerators were designed and costed. The
four incinerators differed in the degree to which heat was
recovered (0 percent, 35 percent, 50 percent, and 70 percent).
A.5 CALCULATION OF CONTROLLED EMISSIONS .
Both the flare and the incinerator were designed to
obtain a 98 percent destruction efficiency. The controlled
emissions would then be the 2 percent of baseline emissions
that remain in the exit stream from the control device. For
the example stream, controlled emissions were calculated as:
100 - Control Efficiency
Controlled Emissions = Baseline Emissions *
100
3.38 Mg
VOC
-
yr
100 - 98
-
100
= 0.0676 Mg
VOC
yr
and
3.38 Mg
HAP
yr J
100 - 98 | HAP
= 0.0676 Mg
yr
100
A-5
-------
A.6 CALCULATION OF EMISSION REDUCTION
Emission reduction was calculated as follows:
Emission Reduction = Baseline Emission * Control Efficiency
IVOC VOC
3.38 Mg (0.98) = 3.31 Mg
yr J yr
and
I HAP HAP
3.38 Mg (0.98) = 3.31 Mg
yr I yr
A.7 CALCULATION OF COSTS
The cost of controlling air emissions of organic HAP's
from process vents depends on the vent flow rate, the emission
rate, and the control device used.
The most cost effective (lowest annual cost per megagram
of HAP removed) device for the example stream was a flare.
The total annual cost of controlling this stream with a flare
was $46,115 per year. Detailed calculations of flare design
and costing were presented in BID Volume IB, Appendix A. Due
to rounding error, the numbers in Volume IB appendices may
differ slightly from the numbers presented in the text and
appendices of Volume 1C.
A.8 CALCULATION OF COST EFFECTIVENESS
Cost effectiveness for the control of HAP emissions from
SOCMI process vents was defined as the total annual control
cost per megagram of HAP emission reduction. For example/
total annual costs ($/yr)
HAP cost
effectiveness =
for flare HAP emission reduction (Mg/yr)
HAP cost
effectiveness
for example
stream
$46,115
3.31 Mg/yr
= $13,932/Mg HAP
A-6
-------
Thus, the HAP cost effectiveness for applying the design flare
to the example stream was approximately $13,900/Mg HAP.
A-7
-------
-------
APPENDIX B
EXAMPLE IMPACTS FOR ADDITION OF A COMBUSTION
DEVICE TO A TRANSFER LOADING RACK
B.I INTRODUCTION
This appendix provides an example calculation of primary
air pollution impacts and cost impacts associated with the
control of organic emissions from a tank truck or tank car
loading rack through the addition of a combustion control
device. The following example calculation represents an
example facility located in Louisiana, which has a required
State control level less than 98 percent. The facility
transfers four materials including, ethylene dichloride,
formaldehyde, methanol, and vinyl chloride. Calculation data
for the facility is presented in Table B-l. This facility is
represented by Tank Car Model Rack Number 6 and Tank Truck
Model Rack Number 18 in Tables 3-5 to 3-8 in the text of
Volume 1C.
Because transfer racks vary in size and transfer various
organic chemicals, it is important to note that the emissions
and cost impacts from these two example racks do not represent
the impacts for all transfer racks in the SOCMI. These racks
were selected only to illustrate the calculations necessary to
estimate the impacts of controlling emissions from transfer
operations in the SOCMI.
The purpose of this appendix is to demonstrate the
approach used in the RON analysis. In the calculations below,
all significant figures have been retained until the final
calculation to make it easier for the reader to follow the
calculation and to avoid potential error due to round off of
intermediate calculations. It should not be inferred that the
B-l
-------
rH
CQ
S
CQ
*•
a
!1
a
0
1—
u
o
"o
c
m
J=
"S
2:
(D
1
1
o
iZ
CU
11
v^
a
cn
CO
CO
CO
CD
0
in
CO
in
a>
o
^.
LO
o
CO
*-^
t_
n
01
g^
^j
3
&
cn
o
JZ
r—
t-
m
o
I—
e
'x
s:
CO
LO
t— 1
CO
CO
CO
CO
in
CO
o
in
CO
01
cn
—
>^
^"~
1^
ZJ
J-
I—
o
H-
c
t—
E
E
'x
to
CO
CD
f^_
CJ
in
CO
to
in
CO
CO
co
*— 4
0
CO
t_
r—
to
cn
§?
*-*
j i
3
CL
JC
cn
o
i—
t-
tO
o
c
to
t—
m
4-1
CO
m
CM
*— <
CD
0
CJ
in
CO
CO
CD
r-4
CO
cn
— t
u
^^
to
cn
*— '
4-1
CL
cn
3
O
| —
TV*
CJ
3
£_
f—
tO
tO
3
4-1
U
>
s
o
s.
i.
a
o
as
i —
o
4-
LO
C
O
CO
CO
'i
LU
tu
CO
tO
CO
in
in
W
o
CD
CO
CO
CJ
CD
0
CO
— *
t~
cn
3-
o
»
s^_
cn
^-,
o
Q£.
^
U
t_
1—
_^
C
(O
h-
t-
CO
c
o
to
CO
1
ID
tu
'o
t-
c
o
CJ
CO
CO
CO
0
CD
CO
CO
CD
CD
CD*
CO
r-*
CJ
"t?
•i?
cn
2E
U
o;
to
CJ
c
to
H-
s_
C
o
4-1
u
1
CXL
C
o
CO
CO
ii
1 —
w
o
CD
CO
CO
Cv
o
o
cc
*~
^-*.
u
•C**
c
3_
*"•
t
Q
_i
C
(
h
f
h
r
4-
•i:
c
1
<:
i
3
^
1
•o
CU
to
CO
to
CO
4J
1 — 1
.
to
CL
^
*
"H
C
tO
4-J
tu
to
s
01
CU
t_
3
CO
V)
0)
a.
i-
a
to
to
03
to
o
'E
o
tu
JC
tu
CO
o
CU
_Q
O
tu
N
O
4-J
'm
3
CT
tu
to
(O
o
o
CO
4-1
£
o
O
c
o
CO
CO
"i
(D
1
fO
CU
I-
tu
CL
to
•M
to
tu
( i
O
CO
c
o
CO
CO
4-1
o
tu
to
0)
_ci
4-J
T3
c
fO
tu
u
3
CO
CO
tu
f-
CL
tu
c
3
•a
0)
t-
tu
4-
co
c
to
u
CO
to
u
tu
CJ
CU
-Q
CO
CO
c
to
0
tu
-C
(_
o
CO
o
to
to
'i
tu
tu •
£= tU
.-- t-
r— OJ
tU -C
CO CL
(O CO
-Q O
4- -M
O (0
C 0)
0 JZ
(0 O
3
o to
i — c:
to o
o ••—
to
03 W
tu
-M
O 4-1
OJ 3
4- _C
tO 4J
O
c tu
5-
"O ZJ
•»- to
T) CO
0)
4-1 J_
•*— CL
0) 0)
'r- C
0
r— -a
_c tu
0 t_
r- 0)
C* ^V)
• r- C
4- 4-1
O
tu
t_ J2
tu
4- -0
CO r—
S §
t- 5
"*"* Q)
O ••-
4- t-
o
O- t —
-------
intermediate results represent the actual number of
significant figures.
B.2 MODEL ASSIGNMENT
The model assignment of tank truck and tank car transfer
racks is based on the number of materials produced at a
facility and the total maximum throughput of these materials
to tank cars and tank trucks. The example facility produces
four materials with total tank car and tank truck maximum
throughputs of 33.3 MM gal/yr and 15.3 MM gal/yr,
respectively. Using Table 3-2 of the text, a single 16-arm
rack is required for transfer of materials to tank cars.
Similarly, using Table 3-3 of the text, a single 4-arm rack is
required for transfer of materials to tank trucks.
B.3 CALCULATION OF BASELINE EMISSIONS
To calculate the baseline emissions associated with tank
car and tank truck transfer racks, the uncontrolled emissions
corresponding to each material must first be calculated. This
is done for each vehicle by multiplying the actual throughput
of the material by the emission factor of the material. The
uncontrolled emissions for each material are then reduced by
the quantity (1 - the fractional State level control
efficiency) to.obtain baseline emissions. If a material is
not controlled at the State level, the baseline emissions are
equal to the uncontrolled emissions. The following equations
illustrate the calculation of baseline emissions for the
methanol produced at the example facility.
The tank car and tank truck emission factor for methanol
is calculated from the following equation:
Emission Factor (lb/1000 gal) — 12.46 * S * VP * MW/(T + 460)
where:
S = 0.6 = A saturation factor;
VP = 1.93 = Vapor pressure of methanol (psi);
MW = 32.04 = Molecular weight of methanol (Ib/lb-mol);
and
T = 77 = Loading temperature (°F)
B-3
-------
Emission
Factor
= 12>46 * .6 * 1.93 psi * 32.04
Ib
Therefore:
Tank Truck
Uncontrolled= (6.85<
Emissions
yr
Ib-mol
./537 = .861.
Ib
.861 Ib
lOOOgal
*(453.593
Ib
)*
1000 gal
Mg
106g
= 2.68
Mg
yr
Tank Car
Uncontrolled =(15.616*io6.
Emissions ^r
.861 Ib
lOOOgal
453.593
Ib
Mg
106g
= 6.11
Mg
yr
The example facility is located in Louisiana, which
requires 90 percent control of materials having a vapor
pressure greater than or equal to 1.5 psi and an actual
throughput to either tank trucks or tank cars greater than
40,000 gal/day. The throughput of the chemical in gallons per
day is calculated by dividing the throughput of the chemical
in gallons per year by the number of days per year that the
chemical is transferred, as shown below for methanol transfer
to tank cars at the example facility. The number of days that
a chemical is transferred is represented by a capacity
utilization factor, which represents the fractional part of a
year that the chemical is transferred. For methanol, the
capacity utilization factor is equal to 0.8 meaning the
chemical is transferred on an average 80 percent of the year.
All of the chemicals meet the vapor pressure criteria for
State control. However, the throughput of methanol to the
tank car transfer rack is the only material meeting the
throughput criteria.
Methanol
Tank Car
Throughput
(gal/day)
15.616.
(365 days/yr) * .80
gal/day
B-4
-------
Because the throughput of methanol (53,479 gal/day) exceeds
the State's limit of 40,000 gal/day, emissions of methanol
must be reduced by 90 percent. Therefore:
Baseline
Emissions of
Methanol to the
Tank Car
Transfer Rack
6.11
Mg
yr
1 -
90
100
= 0.611
Mg
yr
The baseline emissions for methanol transfer to the tank truck
transfer rack, and the baseline emissions for ethylene
dichloride transfer to the tank truck and tank car transfer
racks are equal to the respective uncontrolled emissions
because they do not meet the requirements for State control.
The emission factors for formaldehyde and vinyl chloride are
zero because these materials^ have a vapor pressure greater
than 14.7 psi. It is assumed that these two materials are
transferred under pressure and their uncontrolled and baseline
emissions are equal to zero. Although there is a NESHAP for
transfer of vinyl chloride, it did not affect the calculation
of baseline emissions for the HON analysis because it was
assumed that vinyl chloride would be transferred under
pressure with no emissions to the atmosphere.
To obtain the total baseline emissions per transfer rack,
the baseline emissions for each material are summed on a rack
basis. For the example facility, the tank truck rack total
baseline emissions are 4.55 Mg/yr, and the tank car rack total
baseline emissions are 3.46 Mg/yr.
The baseline emissions for transfer of benzene are
calculated a little differently. The benzene transfer NESHAP
requires 98 percent control of benzene if the throughput to
tank cars or tank trucks is greater than 343.5 thousand
gallons per year. Therefore, for all facilities producing and
transferring benzene in large enough quantities, the baseline
emissions equal the uncontrolled emissions reduced by
98 percent.
B-5
-------
Benzene
Baseline
Emissions
Mg/yr
Uncontrolled
Emissions
Mg/yr
* (1 - .98)
B.4 CONTROL DEVICE ASSIGNMENT
Three types of control devices are possibly assigned to
tank truck and tank car transfer racks at SOCMI facilities,
including: flares, incinerators, and incinerator plus
scrubber systems. Since the example facility produces two
halogenated materials (ethylene dichloride and vinyl
chloride), both the tank truck transfer rack and the tank car
transfer rack are assigned an incinerator plus scrubber
system. For those SOCMI facilities producing all
nonhalogenated materials, either a flare or an incinerator is
assigned to the transfer racks based on which device has the
lower cost.
B.5 CALCULATION OF CONTROLLED EMISSIONS
Controlled emissions are calculated on a material basis
and on a rack basis. For all facilities without required
State control or with required State control levels less than
98 percent, the controlled emissions on a material and a rack
basis are estimated as follows:
Controlled Baseline
Emissions = Emissions *
(Mg/yr) (Mg/yr)
-i
Control Efficiency
100
For these two types of facilities, the control efficiency is
98 percent, which represents the control efficiency of flares
and incinerators.
For those SOCMI facilities having State control levels of
98 percent or greater or benzene control levels of 98 percent,
the controlled emissions for each material and each transfer
rack are equal to the baseline emissions. These facilities
are not assigned an additional HON control device, because the
control level is already at least.98 percent.
B-6
-------
The following example illustrates the calculation of the
controlled emissions for methanol at the example facility.
Tank Truck
Controlled
Emissions
(Mg/yr)
= 2.68
Mg
yr
1 -
98
100
= .0536
Tank Car
Controlled
Emissions
(Mg/yr)
= .611
Mg
yr
1 -
98
100
= .0122
The total controlled emissions for each transfer rack are
calculated by summing the transfer rack controlled emissions
for each material. For the example facility, the total
controlled emissions for the tank truck and tank car racks are
.0911 Mg/yr and .0691 Mg/yr, respectively.
B.6 CALCULATION OF EMISSION REDUCTION
Emission reduction is calculated on a material and on a
rack basis. For all materials and transfer racks at all
facilities, the emission reduction is the difference between
the baseline emissions and the controlled emissions.
Therefore, facilities having State control levels of
98 percent or greater or benzene control of 98 percent have
emission reductions equal to zero. The following example
illustrates the emission reduction calculation for methanol at
the example facility.
B-7
-------
Tank Truck
Emission Reduction = 2.68
(Mg/yr)
Mg
yr
- .0536
=2.63
yr
Tank Car Emission
Reduction
(Mg/yr)
= .611
- .0122
yr
= .599
yr
The total emission reduction for each transfer rack is
calculated by summing the transfer rack emission reduction for
each material. For the example facility, the emission
reduction for each chemical is shown in Table B-l and the
total emission reduction for the tank truck and tank car racks
is 4.47 Mg/yr and 3.39 Mg/yr, respectively.
B.7 CALCULATION OF COSTS
The costs of controlling air emissions or organic HAP's
from tank truck and tank car transfer operations depend on the
type of control device assigned to the transfer rack and the
flow rate and concentration of the organic HAP's to the
control device. It is assumed that each transfer rack is
equipped with at least one dedicated control,device.
The tank truck and tank car racks in the example facility
each require an incinerator plus scrubber system with a
control efficiency of 98 percent. The total annual cost of
the control devices for the tank truck and tank car transfer
racks at this facility are $74,321/yr and $84,448/yr,
respectively. Detailed cost analysis calculations for this
facility are provided in BID Volume IB, Appendix B.
B.8 CALCULATION OF COST EFFECTIVENESS
Cost effectiveness for the control of HAP emissions from
tank truck and tank car transfer racks is defined as the total
annual control cost per megagram of HAP emission reduction.
For example,
B-8
-------
HAP Cost Effectiveness
for Transfer Rack
Control Device ($/Mg)
Total Annual Cost ($/yr)
HAP Emission Reduction (Mg/yr)
HAP Cost Effectiveness
for Example Tank
Truck Rack Control
Device ($/Mg)
_ 74,321 ($/yr)
4.47 (Mg/yr)
= $16,627/Mg
HAP Cost Effectiveness
for Example Tank Car
Rack Control Device ($/Mg)
84,448 ($/yr)
= '. v*/y ' = $24,911/Mg
3.39 (Mg/yr)
The cost effectiveness for the tank truck and tank car
transfer rack control devices at the example facility is
approximately $16,700/Mg and $24,900/Mg, respectively.
B-9
-------
-------
APPENDIX C
EXAMPLE IMPACTS FOR APPLICATION OF AN INTERNAL FLOATING ROOF
OR A REFRIGERATED CONDENSER ON A STORAGE TANK
C.I INTRODUCTION
This appendix presents example calculations of primary
air pollution impacts and cost impacts for controlling
emissions from fixed roof storage tanks with either internal
floating roofs or refrigerated condenser systems. The model
tank farms discussed here are identical to model tank farms #7
and #10 presented in Table 4-3 of the text. Table C-l of this
appendix lists design parameters for the two tank farms.
Because storage tanks vary in capacity and store various
organic chemicals, it is important to note that the emissions
and cost impacts from these two example tank farms do not
represent the impacts for all tank farms in the SOCMI. These
tank farms were selected only to illustrate the calculations
necessary to estimate the impacts of controlling storage tank
emissions in the SOCMI.
The purpose of this appendix is to demonstrate the
approach used in the HON analysis. In the calculations below,
all significant figures have been retained until the final
calculation to make it easier for the reader to follow the
calculation and to avoid potential error due to round off of
intermediate calculations. It should not be inferred that the
intermediate results represent the actual number of
significant figures.
C.2 MODEL ASSIGNMENT
The number and size of tanks in the model tank farms were
determined using the annual storage throughput and Table 4-1
of the text. .For model tank farm #7, a throughput of
C-l
-------
TABLE C-l. MODEL TANK DESIGN PARAMETERS
Parameter Description
Units
Model Tank Model Tank
Farm #7 Farm #10
Value; Value
Number of tanks
Tank capacity
Annual throughput per
tank (AN)
Tank orientation
Tank diameter (D)
Tank height
Number of columns
(Nc)
Effective column
diameter (Fc)
Average tank vapor
space height (H)
Adjustment factor for
small diameter tanks
(Cd>
Tank paint factor
(Fp) - white roof and
aluminum color shell
Average ambient
diurnal temperature
change (AT)
Product factor (Kc) -
organic liquid other
than crude oil
gal
gal/yr
feet
feet
feet
feet
dimensionless
dimensionless
Op
dimensionless
10 3
2,000,000 20,000
18,057,775 2,472,727
vertical vertical
fixed roof fixed roof
85
47
1
1.0
23.5
1.0
1.3
20
1.0
15
15
1
NA
7.5
0.7306
1.3
20
1.0
Turnover factor (K^)
Seal factor (Kg) for
liquid-mounted seals
Average wind speed
(V)
Seal related wind
speed exponent (n)
Vapor pressure
function (P*)
dimensionless
Ib mol/[ft
(mi/hr)n yr]
mi/hr
dimensionless
dimensionless
1.0
3.0
10
0
0.0022
0.409
NA
10
NA
NA
C-2
-------
TABLE C-l (Continued).
Parameter Description
Units
Model Tank Model Tank
Farm #7 Farm #10
Value Value
Shell clingage factor
(C)
Fitting loss factor
for controlled deck
fittings (Ff)
Seam loss factor for
bolted decks
Deck seam length
factor (KD)
Stored product - HAP
bbl/1000 ft2
Ib mol/yr
Ib mol/ft-yr
feet
0.0015
381.2
0.34
0.2
NA
NA
NA
NA
styrene vinylidene
chloride
Product molecular
weight (My)
Product specific
gravity (actual)
Product vapor
pressure at 25 °C
Atmospheric pressure
(PA)
Product average
market price
Ib/lb mol
— — —
psia
psia
$/kg
104.16
0.906
0.13
14.7
0.97
96.
1.
11
14
1.
,94
21
.6
.7
57
NA = not applicable
C-3
-------
180.58 MMgal/yr requires ten 2,000,000-gallon tanks. For
model tank farm #10, a throughput of 7.42 MMgal/yr requires
three 20,000-gallon tanks.
Selection of model tank type is based on the baseline
control requirements imposed by applicable State and federal
regulations and the chemical properties of the stored chemical
(vapor pressure, compatibility with aluminum, and whether or
not the chemical is halogenated). For both example model tank
farms, there were no baseline control requirements and no
restrictions due to chemical properties, so tank farms #7 and
#10 comprise fixed roof tanks.
C.3 CALCULATION OF BASELINE EMISSIONS
Baseline emissions of HAPs from the two model tank farms
are calculated as the sum of breathing and working losses from
uncontrolled fixed roof tanks. Uncontrolled breathing and
working losses are estimated using AP-42 emission equations.
HAP baseline emissions from model tank farm #7 are 53.1 Mg/yr
(5.31 Mg/yr per tank). From model tank farm #1.0, HAP baseline
emissions are 46.8 Mg/yr (15.6 Mg/yr per tank). The following
equations illustrate the estimation of baseline emissions from
a single tank in model tank farm #7.
Baseline emissions = Breathing losses + Working losses
where:
Breathing
Loss
Mg
yr
= 1.02 * 10
-5
*v
PA - P
0.68
Dl.73H0.51AT0
where:
Mv
PA
p
D
H
104.16 = molecular weight of vapor in storage
tank (Ib/lb mole),
14.7 = average atmospheric pressure (psia),
0.13 = true vapor pressure at liquid storage
temperature of 25 °C (psia),
85 = tank diameter (ft),
23.5 = average vapor space height (ft),
C-4
-------
AT = 20 = average ambient diurnal temperature change
(OF),'
Fp = 1.3 = paint factor (dimesionless),
Cd = 1.0 = adjustment'factor for small diameter
tanks (dimensionless),
Kc = 1 = product factor (dimensionless).
Breathing losses = 1.025 x 10~5 * 104.16 * ( °-12748 _\0.68
\14.7 - 0.12748/
* (85)1-73'* (23.5)°-51 * (20)°-50
* (1.3) * 1.0 * 1.0
Breathing losses = 2.70 Mg/yr per tank
Working Loss (Mg/yr) = 1. 089 x lO'8 MV(P) (AN) (KN) (Kc)
where:
AN ^..1 18,057,775 = total throughput per year (gal),
replaces N x V in AP-42 eguation, where:
N = number of turnovers per year,
V = tank capacity (gal),
KJJ = 1 = turnover factor (dimensionless) ,
My, P, and KC as defined above.
Working losses = 1.089 X 10 ~8 * 104.16 * 0.12748
* 18,057,775 * 1.0 * 1.0
Working losses = 2.61 Mg/yr per tank
Therefore:
Baseline emissions =2.70 Mg/yr +2.61 Mg/yr
Baseline emissions = 5.31 Mg/yr per tank
For the entire tank farm (10 tanks), baseline HAP emissions
will be 53.1 Mg/yr.
C.4 CONTROL DEVICE ASSIGNMENT
Two control technologies were evaluated for storage tanks
in the HON analysis—tank improvements (i.e., installing an
internal floating roof inside a fixed roof tank or upgrading
an existing internal floating roof) and refrigerated
C-5
-------
condensers. Baseline control technologies and chemical
properties were the major factors in selecting the control
technology appropriate for a particular model tank.
Because neither of the example model tank farms had a
control device in the baseline, chemical properties determined
the assignment of a HON control device. In general, an
internal floating roof having a primary, liquid-mounted seal
and controlled deck fittings was assigned to a fixed roof
tank. This was the case for model tank farm #7.
Refrigerated condensers were assigned to control tanks
storing halogenated compounds and some glycol ethers since
these chemicals have been found to be incompatible with the
aluminum used to construct most internal floating roofs. A
control efficiency of 95 percent was selected for the
condensers because it is equivalent to the emission reduction
achievable using an internal floating roof. Because model
tank farm #10 stores vinylidene chloride (a halogenated
material), refrigerated condensation was assigned as the
control technology.
C.5 CALCULATION OF CONTROLLED EMISSIONS
C.5.1 Internal Floating Roof Tanks
HAP emissions from a storage tank controlled with an
internal floating roof are calculated as the sum of the
withdrawal losses, rim seal losses, deck fitting losses, and
seam losses. These emissions are estimated using AP-42
emission equations. Controlled HAP emissions from a single
tank in model tank farm #7 are estimated as follows:
Controlled
HAP emissions
from an
internal floating
roof tank
Withdrawal
losses
Rim seal
losses
(primary
liquid-
mounted
seal)
Controlled Seam
deck losses
fitting (bolted
losses deck)
(a) First, calculate withdrawal losses:
C-6
-------
Withdrawal Loss (Mg/yr) = 1-018 *10"5 QCWL
N
where :
Q
C
N
= 18,057,775 = throughput (gal/year),
= 0.0015 = shell clingage factor (bbl/1,000 ft2),
•= 7.5605 = average organic liquid density (Ib/gal),
= 85 = tank diameter (ft),
= 1 = number of columns (dimensionless),
= 1 = effective column diameter (ft) [column
perimeter (ft)/pi],
.thdrawa]
Loss
(Mg/yr)
1.018 * 10~5 * 18,057,775 * 0.0015 * 7.5605
85
1 +
1*1
85
Withdrawal Loss = 0.025 Mg/yr
(b) Next, calculate rim seal losses:
Rim Seal Loss (Mg/yr) =
2204.6
where:
Mv
PA
p
V
n
104.16 = molecular weight of vapor in storage
tank (Ib/lb mole),
14.7 = average atmospheric pressure (psia),
0.13 = true vapor pressure at liquid storage
temperature of 25 °C (psia),
85 = tank diameter (ft),
1 = product factor (dimensionless),
3 = seal factor [lb-mole/(ft (mi/hr)n yr)],
10 = average wind speed at tank site (mi/hr),
0 = seal related wind speed exponent
(dimensionless) ,
0.0022 = vapor pressure function
(dimensionless),
C-7
-------
0.13
p* =
PA - 14-7
i +
-L P
PA.
0.5
2
1 +
0.13
14.7
0.5
2 = 0.0022
Rim Seal Loss (Mg/yr) =
3 * 10° * 0.0022 * 85 * 104.16 * 1
2204.6
Rim Seal Loss = 0.026 Mg/yr
(c) Next, calculate fitting losses:
Fitting Loss (Mg/yr) = Ffp MVKC
2204.6
where:
Ff
P*, Mv/ K
fitting loss factor (Ib-mole/yr),
0.0228D2 + 0.79D + 147.9
0.023 (85)2 + 0.79 (85) + 147.9
381.2
as defined above.
Fitting Loss (Mg/yr) =
381.2 * 0.0022 * 104.16 * 1
2204.6
Fitting Loss = 0.039 Mg/yr
(d) Calculate deck seam losses:
, . KnSnD2P*MK
Deck Seam Loss (Mg/yr) = D D
2204.6
where :
D, P*, Mv/ Kc
= 0.34 = deck seam loss per unit seal length
factor (Ib-mole/ft yr),
= 0.2 = deck seam length factor (ft/ft2),
= as defined above.
C-8
-------
Deck Seam Loss (Mg/yr) =
0.34 * 0.2 * 852 * 0.0022 * 104.16 * 1
2204.6
Deck Seam Loss = 0.051 Mg/yr
(e) Finally, calculate controlled emissions for the
example tank as the sum of withdrawal loss, rim seal loss,
fitting loss, and deck seam loss.
Controlled
HAP emissions
from an internal
floating roof
tank
= 0.025 Mg/yr + 0.026 Mg/yr + 0.039 Mg/yr + 0.051 Mg/}
Controlled HAP emissions
from an internal =0.14 Mg/yr
floating roof tank
For the entire tank farm (10 tanks), controlled HAP emissions
will be 1.41 Mg/yr.
C.5.2 Refrigerated Condenser Systems
Controlled HAP emissions from a storage tank with a
refrigerated condenser are based on the tank's baseline HAP
emissions and the removal efficiency of the condenser
(95 percent). Controlled HAP emissions from a single tank in
model tank farm #10 are estimated as follows:
Controlled
HAP emissions
Baseline
emissions
Condenser
1 - removal
efficiency
Controlled HAP emissions = 15.59. * (1 - 0.95)
Controlled HAP emissions = 0.78 Mg/yr
For the entire tank farm (3 tanks), controlled HAP emissions
will be 2.34 Mg/yr.
C-9
-------
C.6 CALCULATION OF EMISSION REDUCTION
HAP emission reductions from storage tanks are calculated
as the difference between baseline emissions and controlled
emissions. HAP emission reductions from model tank farms #7
and #10 are 51.7 Mg/yr (5.17 Mg/yr per tank) and 44.5 Mg/yr
(14.8 Mg/yr per tank), respectively.
C.7 CALCULATION OF COSTS
The costs of controlling air emissions of organic HAP's
from storage vessels depend upon the emission rate from the
vessel and the specific control device used. Some cost
savings could be achieved at larger facilities if controls
were centralized (e.g., all tanks in one tank farm vented to
the same refrigerated condenser). However, for the HON
analysis, it was assumed that each individual tank would be
equipped with a dedicated control device.
To control each of the 10 tanks in model tank farm #7
with an internal floating roof having a liquid-mounted primary
seal and controlled deck fittings, the total annual cost would
be $113,368/yr or $11,337/tank-yr. To control each of the
3 tanks in model tank farm #10 with a refrigerated condenser
achieving an emission reduction of 95 percent, the total
annual cost would be $116,463/yr or $38,821/tank-yr. Detailed
calculations of control costs for model tank farms #7 and #10
are presented in BID Volume IB, Appendices E and C,
respectively. Due to rounding error, the numbers in Volume IB
appendices may differ slightly from the numbers presented in
the text and appendices of Volume 1C.
C.8 CALCULATION OF COST EFFECTIVENESS
Cost effectiveness for the control of HAP emissions from
SOCMI storage tanks is defined as the total annual control
cost per megagram (Mg) of HAP emissions reduction. For
example,
HAP cost
effectiveness
for model tank
farm #7 ($/Mg)
Total annual cost ($/yr)
HAP emissions reduction (Mg/yr)
C-10
-------
HAP-cost
effectiveness
for model tank
farm #7
$113,368/yr
51.71 Mg/yr
HAP cost
e f f e ct i venes s
for model tank
farm #7
= $2192/Mg of HAP reduced
Thus, HAP cost effectiveness for applying the'design internal
floating roofs in model tank farm #7 is approximately
$2200/Mg. HAP cost effectiveness for applying the design
refrigerated condenser to model tank farm #10 is approximately
$2600/Mg.
C-ll
-------
-------
APPENDIX D
EXAMPLE IMPACTS FOR APPLICATION OF STEAM STRIPPING
TO CONTROL HAP AND VOC EMISSIONS FROM, WASTEWATER
D.1 INTRODUCTION
This appendix presents an example calculation of primary
air pollution impacts and cost impacts for applying a steam
stripper to model wastewater stream No. 63 described in
Table 5-4 of the text. The model stream selected is
characterized by an individual stream flow of 100 £pm, a HAP
concentration of 1600 mg/£, a fraction emitted factor of 0.25,
and a strippability factor of 0.70. In addition, for the
purpose of estimating control costs, the assumed flow of
combined streams requiring treatment is 500 £pm. It should be
noted that facilities using steam stripping to remove organic
compounds from wastewater will likely not apply a separate
steam stripper to each individual wastewater stream; instead,
facilities will more likely combine wastewater streams
whenever possible for more economical treatment. Therefore,
cost impacts of steam stripping are dependent on the combined
stream flow rates.
Because wastewater streams can vary greatly in flow,
composition, and HAP volatility, it is important to note that
the emissions and cost impacts for this model stream do not
represent the impacts for all streams in the SOCMI. It should
also be noted that this model stream does not contain benzene
or vinyl chloride. This model stream was selected solely to
present example calculations of impacts resulting from
controlling wastewater streams in the"SOCMI.
D-l
-------
D.2 MODEL ASSIGNMENT
Model wastewater streams were developed from data
received from Section 114 wastewater questionnaires sent to
nine corporations in the SOCMI solely for the purpose of
presenting the range of impacts associated with wastewater
streams. Each of the 84 model streams are defined by flow
rate (£pm/Gg/yr), HAP concentration (mg/£), and volatility.
Model stream development is discussed in detail in
Section 6.2.2 of this document.
Where actual wastewater stream data were obtained, these
data were used directly for estimation of baseline emissions
and control impacts for SOCMI product processes. However, for
product processes where no actual stream data were available,
wastewater flow rate, HAP concentration, and volatility were
estimated using (1) available information on HAP's used or
produced in the process and the solubility and volatility of
these compounds, (2) the equipment components used in the
process which could be a source of wastewater, and
(3) engineering judgment.
D.3 CALCULATION OF UNCONTROLLED EMISSIONS
Uncontrolled HAP and VOC emissions from SOCMI wastewater
streams are a function of the stream's flow rate, HAP
concentration, and the volatility of compounds present in the
stream. The uncontrolled HAP and VOC emissions from the
example model stream are 21.0 Mg/yr and 77.4 Mg/yr,
respectively. Uncontrolled HAP emissions (Mg/yr) from this
model stream are based on a stream flow rate of 100 £pm and a
HAP concentration of 1600 mg/£.
Uncontrolled
HAP Emissions = Fe
Wastewater
Stream Flow
Rate
Wastewater
Stream HAP
Concentration
Where:
Fe
The fraction of HAP mass emissions to the air
per total HAP mass loading in the wastewater
stream.
D-2
-------
Uncontrolled
HAP Emissions
(from the example
model stream)
= 0.25 * 100
10
-9
£ * 1600 mg *
min £
Mg * 60 min * 8760 hr
mg hr yr
Uncontro11ed
HAP Emissions
(from the example
model stream)
=21.0 Mg/yr
Uncontrolled VOC emissions from SOCMI wastewater streams
are calculated in a similar manner with one exception. The
uncontrolled VOC emissions calculation includes a VOC/HAP
emissions ratio that was developed from Section 114
questionnaire data. This ratio allows the calculation of
uncontrolled VOC emissions based on the uncontrolled HAP
emissions.
Uncontrolled Wastewater Wastewater
VOC Emissions = Fe * Stream Flow * Stream HAP * VOC Emissions
Rate Concentration HAP Ratio
where:
VOC Emissions
HAP Ratio
Ratio of total volatile organic
compounds
that would be emitted from
wastewater to the total organic
HAP's that would be emitted from
wastewater. This ratio was
developed from Section 114
questionnaire data and is equal
to 3.68.
Uncontrolled
VOC Emissions
(from the example
model stream)
= 0.25 * 100
— £— * 1600 nig * 3.68 *
min £
10-9 Mg * 60 min * 8760 ii£
mg hr yr
Uncontrolled
VOC Emissions
(from the example
model stream)
=77.4 Mg/yr
D-3
-------
D.4 CALCULATION OF BASELINE EMISSIONS
Because no benzene or vinyl chloride is present in model
stream No. 63, there are no Federal or State wastewater
emission control requirements affecting this model stream.
Therefore, wastewater uncontrolled emissions are also
wastewater baseline emissions. The HAP and VOC baseline
emissions from the example model wastewater stream are
21.0 Mg/yr and 77.4 Mg/yr, respectively.
D.5 CONTROL DEVICE ASSIGNMENT
Many types of control devices are suitable for
controlling volatile organic emissions from wastewater, but
the most universally applicable control technology within the
SOCMI is steam stripping. To estimate impacts of the HON, it
was assumed wastewater streams would be controlled by steam
stripping. It is assumed that facilities will combine
wastewater streams where possible for treatment, in common
steam stripping units.
D.6 CALCULATION OF EMISSION REDUCTIONS AND CONTROLLED
EMISSIONS
Emission reductions of HAP's and VOC's from wastewater
streams are a function of the stream's flow rate, HAP
concentration, and strippability. The HAP and VOC emission
reductions from steam stripping the example model wastewater
stream are 14.7 Mg/yr, and 54.2 Mg/yr, respectively. The HAP
emission reduction is based on the uncontrolled HAP emissions
of 21.0 Mg/yr and a dimensionless strippability factor (Fr) of
0.70.
HAP Emission
Reduction
Uncontro11ed
= HAP Emissions * Fr
where:
Fr
Strippability - The predicted fractional
reduction in emissions achieved by steam
stripping. The strippability predicted for the
example model wastewater stream is 0.70.
D-4
-------
HAP Emission
Reduction (from
the example
model stream)
HAP Emission
Reduction (from
the example
model stream)
= 21.0 Mg * 0.70
yr
= 14.7 Mg/yr
The VOC emission reduction from .SOCMI wastewater streams
is calculated in a similar manner with one exception. The VOC
emission reduction is based on the uncontrolled VOC emissions
of 77.4 Mg/yr.
VOC Emission
Reduction
VOC Emission
Reduction (from
the example
model stream)
VOC Emission
Reduction (from
the example
model stream)
Uncontrolled
VOC Emissions * 0.70
77.4 Mg * 0.70
yr
54.2 Mg/yr
Controlled HAP and VOC emissions from wastewater streams
are calculated as the difference between baseline emissions
and emission reductions resulting from stream stripping. The
controlled HAP and VOC emissions from the example model
wastewater stream are 6.3 and 23.2 Mg/yr, respectively.
Controlled
HAP Emissions
HAP Baseline
Emissions
HAP Emission
Reduction
Controlled
HAP Emissions
(from the example
model stream)
Controlled
HAP Emissions
(from the example
model stream)
21.0 Mg/yr
6.3 Mg/yr
14.7 Mg/yr
D-5
-------
Controlled
VOC Emissions
Controlled
VOC Emissions
(from the example
model stream)
Controlled
VOC Emissions
(from the example
model stream)
VOC Baseline
Emissions
77.4 Mg/yr
23.2 Mg/yr
VOC Emission
Reduction
54.2 Mg/yr
D.7 CALCULATION OF COST
Steam stripping control costs are a function of the total
treated wastewater flow rate. It is assumed that facilities
will combine wastewater streams for treatment in a single
steam stripping unit. As noted in Section 5.2.3 of this
volume, the combined facility flow rate is 500 £pm, and the
individual stream flow rate is 100 £pm for model stream
No. 63. The total annual cost for the steam stripping unit is
$418,000/yr. The average cost for a single stream is
$83,600/yr. Detailed steam stripping control costs for model
stream No. 63 are presented in Appendix D of BID Volume IB.
D.8 CALCULATION OF COST EFFECTIVENESS
Cost effectiveness for the control of HAP emissions from
SOCMI wastewater streams is defined as the total annual
control cost per megagram of HAP emission reduction. The HAP
cost effectiveness for the example model wastewater stream is
rounded off as $5,700/Mg of HAP reduced.
HAP Cost
Effectiveness =
Wastewater Stream's
Total Annual Control Costs
HAP Emission Reduction
HAP Cost
Effectiveness
(for example, model
wastewater stream)
HAP Cost
Effectiveness
(for example, model
wastewater stream)
$83.600/vr
14.7 Mg of HAP Reduced/yr
$5,700/Mg of HAP Reduced
D-6
-------
APPENDIX E
EXAMPLE IMPACTS FOR CONTROL OF VOC AND
HAP EMISSIONS FROM EQUIPMENT LEAKS
E.1 INTRODUCTION
The purpose of this appendix is to demonstrate the
approach used in the HON analysis. The results of many
calculations are shown to one decimal place to make it easier
for the reader to follow the calculation. It should not be
inferred that the results presented here represent the actual
number of significant figures.
The following calculations present the methodology used
to calculate air pollution impacts and cost impacts for
control of VOC and HAP emissions from equipment leak model
units B and E. These model units have the same equipment
counts, but B is an uncontrolled unit and E is a baseline
controlled unit.
E.2 MODEL ASSIGNMENT
A total of six model units were developed from three sets
of equipment counts and two levels of baseline control.
Equipment counts in the model units are based on data from
SOCMI process units. They represent three levels of
complexity: low, medium, and high. The low complexity model
unit has the fewest equipment components, and the high
complexity model unit the most equipment components. Baseline
control in the model units was designated as either
uncontrolled or controlled.
Model units were assigned to SOCMI process units in the
United States based on a national complexity distribution and
the location of the SOCMI process unit. The complexity
distribution specified that 52 percent of process units have
E-l
-------
low complexity, 33 percent have medium complexity, and
15 percent have high complexity. . The control status of a
process unit was determined by the applicability of any
federal, state, or local equipment leak regulations to the
process unit.
E.3 CALCULATION OF BASELINE EMISSIONS
Both Model Units B and E have equipment counts associated
with medium complexity. Model Unit B is uncontrolled and
Model Unit E is baseline controlled. Emissions from the model
units are estimated by multiplying emission factors by
equipment counts as demonstrated in the following equation:
/ VOC \ _ /Emission\ / Hours of \ /Equipment^
\Emissions; \ Factor / \0perationj \ Count /
The emission factors used for uncontrolled model units
for all equipment types except pressure relief seals and open-
ended lines are the SOCMI average factors. The SOCMI average
factors were first published in the Fugitive Emissions
Additional Information Document (EPA-450/3-82-010) . Emissions
from pressure relief seals and open-ended lines are commonly
controlled even in "uncontrolled" process units. It was
assumed that in uncontrolled process units 75 percent of
pressure relief seals and 100 percent of open ended lines were
controlled to the level described in the SOCMI Control
Technology Guideline (CTG) for fugitive emissions (EPA-450/3-
83-006) . Calculation of the emission factor used for pressure
relief seals in Model Unit B is demonstrated below:
SOCMI CTG Pressure relief seal
0.25 * Average + 0.75 * Controlled = uncontrolled
Factor Factor emission factor
= 0.25 * 0.104
hr
+ 0.75 * 0.0582 -^/source
nr
= 0.0697 -^/source
hr
All emission factors used for baseline controlled model units
are based on control equivalent to that specified in the CTG.
E-2
-------
Table E-l presents the equipment counts for Model Units B and
E, and uncontrolled and CTG controlled emission factors.
Using the VOC emissions equation presented above,
equipment leak emissions for each equipment type are
calculated. This is demonstrated for light liquid valves in
Model Unit B. The uncontrolled emission factor for light
liquid valves is 0.0071 Kg/hr/source. Model Unit B has
1,179 light liquid valves. It is assumed that the equipment
operates for the entire year which is equal to 8,760 hours.
( VOC \ = /O
\Emiss ions; \
.0071
hr. valve/
Kg
* (1,179 valves)
= 73,329 Kg/yr
=73.3 Mg/yr
HAP emissions are estimated as 0.78 times VOC emissions. This
HAP-to-VOC ratio is based on stoichiometric data from
135 product-processes. It does apply to any specific process
unit. Thus, HAP emissions from light liquid valves in Model
Unit B equal:
HAP emissions = (73.3 Mg/yr) * (0.78) =57.2 Mg/yr
VOC and HAP emissions from other equipment types are
calculated using the same approach as demonstrated for light
liquid valves. Table E-2 presents baseline emissions for
Model Units B and E. Total VOC and HAP emissions from Model
Unit B are 167.4 Mg/yr and 130.6 Mg/yr, respectively. Total
VOC and HAP emissions from Model Unit E are 110.8 Mg/yr and
86.4 Mg/yr, respectively.
E.4 CONTROL ASSIGNMENT
The equipment leak negotiated regulation requires a leak
detection and repair program (LDAR) to reduce emissions from
light liquid pumps, gas and light liquid valves, and
connectors. For compressors, pressure relief devices, open-
ended lines, and sample connections, the negotiated regulation
calls for design specifications. Based on the control
requirements specified in the negotiated regulation, maximum
E-3
-------
8
0
1
$
^-* ^^
5L CQ C fl)
"Ho u
j^-P-H H
•H 10 rH 3
•^ C in O 0
P tD -H .p to
ij 6 O \
•£. rH W tO M
So) fa 43
P, TJ U \
G ° 9 j?
& £. J— •
+j
CQ C
Q)
to &
•P Oi m
•H -H ,.
H §
i-H r *
0) W u
T3
O T3
s c
id
0)
O
•H
^
0)
CO
1)
c
Q)
•H
&
W
0>] •* O
o o o
o o o
o o o
CO
vo H eg
in t^ o
o o o
000
000
*$ a\ <-\
H r^ r-
-* H
H
'O '^3
•rl -H
PJ ^f
•rl -H
1-5 r-3
•p >1
J-3 ^
U) tJ^ nj
(0 -H Q)
CD iJ ffi
co
0)
1
(0
^H ^*
CO H
CO CM
O O
0 0
<* "*
CTi H
•<* OO
O O
O O
o in
*
T3 ^3
•H -H
2^ c^.
•H -H
1-3 1-3
•P >i
jC ^
en (0
•H CD
i-3 33
to
H
tO
0)
CO
QJ
g
g
co
in
H
o
CO
CM
0
CM
O
ft
(0
*x^
10
(d
O
CO
rH
O-H
(d -P
er>
•O 3
0) fa
H
rH H
O »E3
-P 0
C CO
o
O Q)
C 43
p-P
(d
•o
to
CO
1
CO
CQ
CU
tn
to
ft
c
o
TJ
0)
•H
•rH
o
0)
in
rH
O
fa
4J
O
O •
co
a) i
rH CO
0) 0
43 in
-P 'd'
1
13 •<
O A
5.
Q) *""
to o
(d EH
43 0
Q) CQ
M .y
(d to
0)
in 1-3
rH
0 .P
O 0)
(3
O
EH 1
U co
&
0)
^
$3
•H
Q)
•H
•rl
O
0)
ft
in
rH
0)
rH
0)
•p
o
.p
rrj
0)
H
rH
o
-P
(3
0
0
•0
0)
in
in
to
0)
o
•H
Q)
T3
-------
CQ
0
H
CQ
H
S
W
H
CO
Si
W
Q
^
co
EH
H
D
W
Q
O
*
H
H
CQ
EH
0) g
. W C f? ^"
0) .p "JlJ CM -H >i
^ -H 0) £j [jj ~^J
g C W -H 2
t> * B C-
« W
0) g
Wf* X-H *•"•*
i-i 5 O JH
Q) ,, "II U -H >,
"^ -H 0) ° W ^
SC 01 ,~J M
1—5 03 "2 ^~*
CQ r|""
« ' c~
a) _p ",!{ o< -H >i
"5 -H (1) S S "Jri
to* C 01 ,:j S^
* « r|^
0) g_
__, CQ C 0 ^
<1) j, "J U -H >i
73 -H '3 ° W ^
Sc co • M
E> <« P S.
n 1
Q)
O
•H
0)
CO
^_}
c
0)
g
a
•H
p
&
w
f^ CM rH
in CM o
CO
CO CO rH
r^ H o
•<*
CO CM rH
in t^ o
rH in
CO CO rH
O CO O
CM r^
-O TJ
•H -H
•H -H
^ >H
-P >(
W &> (0
(0 -H 0)
O r4 «
tfl
0)
rH
(0
o r- o CM r*
H
VO (T> t^ i > >
X! > \ \
en to w w
•H
-------
achievable control technology (MACT) emission factors were
developed. Development of MACT emission factors is discussed
in Chapter 6 of BID Volume 1C.
E.5 CALCULATION OF MACT EMISSIONS
MACT emissions are the same for Model Unit B and Model
Unit E. Using the same approach as for baseline emissions,
MACT emissions are calculated by multiplying the equipment
count for each equipment type by the corresponding MACT
emission factor. Table E-3 presents MACT emission factors and
Model Unit B and E MACT VOC and HAP emissions. Total MACT VOC
and HAP emissions from Model Units B and E are 18.2 Mg/yr and
14.2 Mg/yr, respectively.
E.6 CALCULATION OF EMISSION REDUCTIONS
Emission reductions are calculated by subtracting the
baseline emission from the MACT emission. Model Unit B and E
emission reductions are calculated below:
Emission _ Baseline Mact
Reductions Emissions Emissions
Model Unit B „
Mg Mg Mg
VOC Emission = 167.4 —- - 18.2 —- = 149.2 —- VOC
Reductions
yr
yr
yr
Model Unit B „ „ „
Mcr Mcr Mg
HAP Emission = 130.6 —- -14.2 — = 116.4 —- HAP
Reductions
yr
yr
yr
Model Unit E
MCT Mcf Mcr
HAP Emission = 110.8 — - 18.2 —- = 92.6 —- HAP
Reductions
yr
yr
yr
Model Unit E , „
Ma Mg Mg
HAP Emission = 86.4 — - 14.2 —- =72.2 —- HAP
Reductions
yr
yr
yr
E-6
-------
CO
ss
o
H
co
CO
H
w
EH
u
A
H
Q
rH r, -H g
(J) HH g ^_,
"° "c W
•^ id
CJ
ro o
CQ ^ W
•P rj C *~>
•rH £7 O rH
C:i -iH >,
^T *^
ID 2 w \
*e* W ^i
nH fyi *rH J»J
Q) ^^ € **""*"
o ^ w
g §
w ^.^
i Q)
00°
o td ^
> fc Q
(d c^
U -H ^
^j! jj
w ^
Q)
O
•H
£
CO
-P
0)
a
•H
£3
w
CO CO H
o in o
«* «* H
o t^- o
VO
H r>.
CM H CO
H t^ CM
o o o
o o o
o o o
000
•d -d
•rH -H
&1 &
•H -H
J J
£ >*
03 ^j^ (^
id -H a)
O I-H K
tn
Q)
>
(d
F~ P*
O O
C* C*
O O
CO
CM H
O CM
O O
O O
TJ rrj
•H -rl
& &
•rH -rH
J Hi
1 1 ^^
X! >
&>(d
•H Q)
FH K
w
rH
(d
0)
CQ
ft
g
g
CO O
0 0
^ O
o o
n
CM
CM
o
o o
Jl Jj
o o
o. o.
(d (d
\ \
CQ CQ
(d (d
O O
tn
(d a>
Q) -H
CO rH
cu
rH PH
O
ra a)
tn rH
o 3
M tn tn
ft tn rH
g CD id
O rH CD
O Pi CO
CO O O CM
VO O O -^
rH
O O O CM
CO O O CO
H
in
CO
0
o
o
o o o
rH rH rH
rH rH rH
CQ
C
O
•H
W -P
<0 O CO
C 0) S
•H C O
1-3 C H
O CQ
•d o co
0} H
tn c c w — >
Q) Q) -H rH
CT* | ,—1 = ^ ^
C C ft rtj \
(d Q) g EH &>
rH ft (d 0 B
fe O CQ EH >—
CD
CQ
Q)
-P
g
ft
a)
•d
0)
•P
c
o
c
o
•H
CQ
CQ
3
O
CQ
*d
o
rH
rH
CO
•
VO
O
•H
•P
(!)
w
CQ
0
H
0)
£3
£j *
rH CQ
O rH
> 0
4-)
Q O
H (d
PQ *w
o c
4J 0
•H
rH CQ
Q> CQ
(II __J
*H HI
0) g
Pi Q)
td
E-7
-------
E.7 CALCULATION OF COSTS
Calculation of total annual control costs for Model Units
B and E is done using the same approach as presented in
Appendix F of Volume IB for the hypothetical process unit.
Refer to Appendix F in Volume IB for detailed explanation of
the approach. Table E-4 summarizes the calculation of total
annual costs (not including the recovery credit) for Model
Unit B and E. These are the same costs calculated in
Appendix F of Volume IB. All costs in Table E-4 are based on
monthly valve monitoring.
Recovery credit must be accounted for in order to
complete estimation of total annual costs. Recovery credit is
the cost savings associated with VOC's which, as a result of
implementing control, are not lost to the air through
equipment leaks. It is calculated by multiplying the annual
emission reduction by the average VOC cost of $l,590/Mg.
Annual Emission Reduction * $l,590/Mg = Recovery Credit
The total annual cost including the recovery credit for
each model unit is calculated below.
Model Unit B
Recovery Credit:
149.2 Mg VOC/yr * $l,590/Mg VOC = $237,200/yr
Total Annual Cost:
$169,800/yr - $237,200 = ($67,400) */yr
Model Unit E
Recovery Credit:
92.6 Mg VOC/yr * $l,590/Mg VOC = $147,200
Total Annual Cost:
$163,900 - $147,200 = $16,700
aParenthesis indicate negative cost, i.e., savings.
E-8
-------
TABLE E-4. SUMMARY OF ANNUALIZED CAPITAL COSTS AND
OPERATING EXPENSES FOR MODEL"UNITS B AND Ea
"*
Annual! zed Capital13
Annual! zed LDAR
Monitoring
Repair
Admin, and Support
Annualized Operating
Miscellaneous Annual
Model Unit B
Costs
($)
40,500
45,000
21,700
26,700
21,700
14.200
169,800
Model Unit E
Costs
($)
34,600
45,000
21,700
26,700
21,700
14.200
163,900
aThese values are taken directly from BID Volume IB
Appendix F.
^Capital Costs for Model Unit B are $210,100 for equipment and
$16,300 for initial leak detection and repair. Capital Costs
for Model Unit E are $203,600 for equipment.
E-9
-------
E.8 COST EFFECTIVENESS
Cost effectiveness is calculated for both VOC and HAP
emission reductions using the following equations.
VOC Cost Effectiveness
Total VOC Cost
Annual ^ Emission = Effectiveness
Costs Reduction ($/Mg VOC)
HAP Cost Effectiveness
Total
Annual
Costs
HAP
Emission
Reduction
Cost
E f feet ivenes s
($/Mg HAP)
Cost effectiveness is calculated for Model Units B and E
below:
Model Unit B Cost Effectiveness;
VOC: -$67,400/yr -t- 149.2 Mg VOC/yr = ($450)/Mg VOC
HAP: -$67,400/yr * 116.4 Mg HAP/yr = ($580)/Mg HAP
Model Unit E Cost Effectiveness;
VOC: $16,700/yr •*• 92.6 Mg VOC/yr = $180/Mg VOC
HAP: $16,700/yr * 72.2 Mg HAP/yr = $230/Mg HAP
E-10
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-453/D-92-016C
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE Hazardous Air Pollutant Emissions from
Process Units in the Synthetic Organic Chemical
Manufacturing Industry—Background Information for
Proposed Standards
Volume 1C: 'Model Emission Sources
5. REPORT DATE.
November 1992
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68D10117
12. SPONSORING AGENCY NAME AND ADDRESS
Director, Office of Air Quality Planning and Standards
Office of Air 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
A draft rule for the regulation of emissions of organic hazardous air pollutants
(HAP's) from chemical processes of the synthetic organic chemical manufacturing
industry (SOCMI) is being proposed under the authority of Sections 112, 114, 116,
and 301 of the Clean Air Act, as amended in 1990. This volume of the Background
Information Document presents model emission sources that were developed to
evaluate the national impacts of the proposed rule.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air pollution
Pollution control
SOCMI
Hazardous air pollutant
Storage tank
Transfer rack
Wastewater
Process Vents
Equipment Leaks
Air pollution control
8. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report/
21. NO. OF PAGES
236
20. SECURITY CLASS (This page!
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
------- | | |