EPA
625/6-86-014
United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park, NC 27711
               Technology Transfer
                            EPA/625/6-86/014'
  v>EPA      Handbook
               Control Technologies for
               Hazardous Air Pollutants
                                              cNVIRONMENTAI
                                               PROTECTION
                                                AGENCY
                                              DALLAS, TEXAp

                                                UBUfif

-------
                                    EPA/625/6-86/014
                                       September 1986
          Handbook

Control Technologies for
Hazardous Air Pollutants
 Air and Energy Engineering Research Laboratory
     Office of Research and Development
    U.S. Environmental Protection Agency
      Research Triangle Park, NC 27711
 Center for Environmental Research Information
     Office of Research and Development
    U.S. Environmental Protection Agency
          Cincinnati, OH 45268

-------
                              Notice

This document has been reviewed in accordance with the U.S. Environmen-
tal Protection Agency's peer and administrative  review policies and ap-
proved for publication. Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.

-------
                             Foreword

Today's rapidly changing industrial technologies, products, and practices
frequently carry with them an increasing generation of hazardous air pollu-
tants (HAPs). When energy and material resources are extracted, processed,
converted, and used, the consequent impacts on health and the environment
require that efficient pollution control methods be used.
                                                              /
The 1970 amendments to the Clean Air Act require the U.S. Environmental
Protection Agency (EPA)  to set National Ambient Air Quality Standards for
criteria air  pollutants found throughout  the  country. The Clean Air Act
Amendments also require EPA to review  and  regulate hazardous air pollu-
tants, defined as those air pollutants that can contribute to increased mortal-
ity or serious illness but which are not already regulated as criteria pollutants.

Since the definition of a HAP is very broad and encompasses thousands of
specific compounds, it is  not practical to develop an all-inclusive list of HAP
compounds  and compound-specific control techniques. However, the num-
ber of generic air pollution control techniques available is small, and the
factors affecting the cost and performance of these controls as  applied  to
many noncriteria pollutants have been identified  and discussed in the literature.

The purpose of this handbook is to help  EPA  regional. State, and local air
pollution control  agency  technical  personnel select, evaluate, and cost air
pollution control  techniques for reducing or  eliminating  the emission  of
potentially hazardous air pollutants from industrial/commercial sources. The
information  provided by  this document will be useful for reviewing permit
applications or for informing interested parties  of the type, basic design, and
cost of available HAP control systems.

-------
                       Acknowledgments

This handbook has been adapted from a two-volume EPA report titled Evalu-
ation of Control Technologies for Hazardous Air Pollutants. The report was
written by Robert Y. Purcell, Pacific Environmental Services Inc., Durham,
North Carolina, and Gunseli Sagun Shareef, Radian Corporation, Research
Triangle Park, North Carolina. They were assisted by: Vishnu S. Katari, Karin
C. C. Gschwandtner, Michael K. Sink, and Charlotte R. Clark of Pacific Envi-
ronmental Services Inc.; and Andrew J. Miles, D. Blake Bath, and Glynda E.
Wilkins of the Radian Corporation.  Dr. Bruce A. Tichenor of EPA's Air and
Energy Engineering  Research  Laboratory,  Research Triangle Park, North
Carolina, was the EPA project officer for the  report.

This handbook was adapted and produced by JACA Corporation, Fort Wash-
ington, Pennsylvania. Norman  Kulujian of EPA's Center for Environmental
Research Information, Cincinnati, Ohio, was the EPA project officer for this
publication.

If you wish to obtain the more detailed two volume report, is is available from
the National Technical Information Service (NTIS), 5285 Port Royal Road,
Springfield, Virginia 22161; (703) 487-4650. The order numbers are:

    Volume I. Technical Report; PB 86-167 020; EPA/600/7-86/009a
    Volume II. Appendices; PB 86-1677 038; EPA/600/7-86/009b
                                IV

-------
                             Contents

Nomenclature  	   x
1 Introduction	   1
  1.1   Objective	   1
  1.2   How to Use the Handbook	   2
  1.3   References	   2

2 HAP Emissions and Their Key Physical Properties 	   5
  2.1   Background	   5
  2.2   Identification of Potential HAP's and Emission Sources	•  5
  2.3   Identification of Key Emission Stream Properties	  18
  2.4   References	  19

3 Control Device Selection 	  23
  3.1   Background	  23
  3.2   Vapor Emissions Control	  23
  3.3   Particulate Emissions Control	  33
  3.4   References	  45

4 HAP Control Techniques 	  47
  4.1   Background	  47
  4.2   Thermal Incineration	  47
  4.3   Catalytic Incineration	  53
  4.4   Flares	  60
  4.5   Boilers/Process Heaters	  63
  4.6   Carbon Adsorption	  63
  4.7   Absorption	  69
  4.8   Condensation	  75
  4.9   Fabric Filters	  80
  4.10  Electrostatic Precipitators	  87
  4.11  Venturi Scrubbers	  90
  4.12  References	  94

5 Cost Estimation  Procedure 	  97
  5.1   Objective	  97
  5.2   Total Capital Cost	  97
  5.3   Annualized Operating Costs	107
  5.4   References	 111
Appendices
  A.1   Potential HAP's for Solvent Usage Operations	 113

  B.1   Gas Stream Parameters Calculations	 115
  B.2   Dilution Air Requirements Calculations	 119

  C.1   HAP Emission  Stream Data Form	 121
  C.2   Calculation Sheet for Dilution Air Requirements	 123
  C.3   Calculation Sheet for Thermal Incineration	 124
  C.4   Calculation Sheet for Catalytic Incineration	 130
  C.5   Calculation Sheet for Flares	 137
  C.6   Calculation Sheet for Carbon Adsorption	 141
  C.7   Calculation Sheet for Absorption	 147
  C.8   Calculation Sheet for Condensation	 154
  C.9   Calculation Sheet for Fabric Filters	 159
  C.10  Calculation Sheet for Electrostatic Precipitators	 162
  C.11  Calculation Sheet for Venturi Scrubbers	 164
  C.12  Capital and Annualized Cost Calculation Worksheet	 167

-------
                              Tables


2-1    Potential HAP's and Emission Sources for Solvent Usage
      Operations	   7
2-2    Potential HAP's and Emission Sources for Metallurgical
      Industries	   9
2-3    Emission Sources for the SOCMI	  10
2-4    Potential HAP's and Emission Sources for the Inorganic
      Chemical Manufacturing Industry	  12
2-5    Potential HAP's and Emission Sources for the Chemical
      Products Industry	  14
2-6    Potential HAP's and Emission Sources for the Mineral
      Products Industry	  15
2-7    Potential HAP's and Emission Sources for the Wood
      Products Industry	  16
2-8    Potential HAP's for Petroleum Related Industries	  16
2-9    Potential HAP's for Petroleum Refining Industries	  17
2-10  Emission Sources for Petroleum Related Industries	  17
2-11  Potential HAP's and Emission Sources for Combustion
      Sources	  18
2-12  Key Properties for Organic Vapor Emissions	  19
2-13  Key Properties for Inorganic Vapor Emissions	  19
2-14  Key Properties for Particulate Emissions	  19

3-1    Key Emission Stream and HAP Characteristics for Selecting
      Control Techniques for Organic Vapors from Point Sources	  24
3-2    Current Control Methods for Various Inorganic Vapors	  29
3-3    Range of Capture Velocities	  30
3-4    Summary of Control Effectiveness for Controlling
      Organic Area Fugitive Emission Sources	  31
3-5    Key Characteristics for Particulate Emission Streams	  35
3-6    Advantages and Disadvantages of Particulate Control
      Devices	  40
3-7    Control Technology Applications for Transfer
      and Conveying Sources	  42
3-8    Control Technology Applications for Loading
      and Unloading Operations	  43
3-9    Control Technology Applications for Plant Roads	  43
3-10  Control Technology Applications for Open Storage Piles	  44
3-11  Control Technology Applications for Waste Disposal Sites	  45

4-1    Thermal Incinerator System Design Variables	  49
4-2    Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Thermal Incineration	  53
4-3    Catalytic Incinerator System Design Variables	  55
4-4    Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Catalytic Incineration	  59
4-5    Flare Gas Exit Velocities	  62
4-6    Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Flares	  63
4-7    Carbon Adsorber System Design Variables	  65
4-8    Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Carbon Adsorption	  69
4-9    Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Absorption	  75
4-10  Coolant Selection	  77

                                  vi

-------
                          Tables (continued)
4-11  Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Condensation	  79
4-12  Characteristics of Several Fibers Used in Fabric Filtration	  82
4-13  Comparisons of Fabric Filter Bag Cleaning Methods	  83
4-14  Recommended Air-to-Cloth Ratios for Various Dusts
      and Fumes by Cleaning Method	  85
4-15  Factors to Obtain Gross Cloth Area from Net Cloth Area	  86
4-16  Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Fabric Filters	  86
4-17  Typical Values for Drift Velocity for Various Particulate
      Matter Applications	  88
4-18  Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for ESP's	  89
4-19  Pressure Drops for Typical Venturi Scrubber Applications	  92
4-20  Construction Materials for Typical Venturi Scrubber
      Applications	  93
4-21  Comparison of Calculated Values and Values Supplied
      by the Permit Applicant for Venturi Scrubbers	  93

5-1    Identification of Design Parameters and Cost Curves
      for Major Equipment	  97
5-2    Chemical Engineering Fabricated Equipment Cost Indices	  98
5-3    Unit Costs for Various Materials	  99
5-4    Price of Packing for Absorber Systems	100
5-5    Bag Prices	101
5-6    Identification of Design Parameters and Cost Curves
      for Auxiliary Equipment	103
5-7    Assumed Pressure Drops Across Various Components	105
5-8    Capital Cost Elements and Factors	107
5-9    Unit Costs to Calculate Annualized Cost	108
5-10  Utility/Replacement Operating Costs for HAP Control
      Techniques	109
5-11  Additional Utility Requirements	110
5-12  Estimated Labor Hours Per Shift and Average Equipment
      Life	110

-------
                             Figures
1-1    Steps Used When Responding to Inquiries	  3
1-2    Steps Used When Reviewing Permits	  4

2-1    A Partially Completed HAP Emission Stream Data Form
      for One of Six HAP Emission Streams (#1) Generated at
      a Fictitious Company	  6
2-2    Potential Emission Points for a Vacuum Distillation Column
      Using Steam Jet Ejectors with Barometric Condenser	  10

3-1    Percent Reduction Ranges for Add-on Control Devices	  24
3-2    Effluent Characteristics for Emission Stream #1	  33
3-3    Effluent Characteristics for Emission Stream #2	  34
3-4    Effluent Characteristics for Emission Stream #3	  35
3-5    Effluent Characteristics for Emission Stream #4	  36
3-6    Effluent Characteristics for Emission Stream #5	  37
3-7    Effluent Characteristics for Emission Stream #6	  39
3-8    Effluent Characteristics for a Municipal Incinerator Emission
      Stream	  41

4-1    Schematic Diagram of a Thermal Incinerator System	  48
4-2    Supplementary Heat Requirement vs. Emission Stream Heat
      Content (Dilute Stream/No Combustion Air)	  50
4-3    Supplementary Heat Requirement vs. Emission Stream
      Heat Content (No Oxygen in Emission Stream/Maximum
      Combustion Air)	  50
4-4   Heat Exchanger Size vs. Emission Stream Flow Rate
      (Dilute Stream/No Combustion Air)	  52
4-5    Heat Exchanger Size vs. Emission Stream Heat Content (No
      Oxygen in Emission Stream/Maximum Combustion Air)	  52
4-6    Schematic Diagram of a Catalytic Incinerator System	  53
4-7    Supplementary Heat Requirement vs. Emission Stream Heat
      Content (Dilute Stream/No Combustion Air)	  56
4-8    Supplementary Heat Requirement vs. Emission Stream
      Heat Content (No Oxygen in Emission Stream/Maximum
      Combustion Air)	  57
4-9    Heat Exchanger Size vs. Emission Stream Heat Content	  59
4-10  A Typical Steam-Assisted Flare System	  60
4-11  Adsorption Isotherms for Toluene/Activated Carbon System	  63
4-12  A Typical Fixed-Bed Carbon Adsorption System	  64
4-13  Carbon Requirement vs. HAP Inlet Concentration	  67
4-14  Steam Requirement vs. Carbon Requirement	  68
4-15  A Typical Countercurrent Packed Column Absorber System	  70
4-16  Correlation for Flooding  Rate in Randomly Packed Towers	  72
4-17  NOG for Absorption Columns with Constant Absorption
      Factor AF	  73
4-18  Flow Diagram for a Typical Condensation System with
      Refrigeration	  75
4-19  Vapor Pressure-Temperature Relationship	  76
4-20  Venturi Scrubber Collection Efficiencies	  91
4-21  Psychrometric Chart, Temperature  Range 0-500°F,
      29.92 in Hg Pressure	  94

5-1    Prices for Thermal Incinerators, Including Fan and Motor, and
      Instrumentation and Control Costs	  98

-------
                          Figures (continued)
5-3   Prices for Catalytic Incinerators, Less Catalyst	 99
5-4   Prices for Carbon Adsorber Packages	 99
5-5   Prices for Custom Carbon Adsorbers,  Less Carbon	100
5-6   Prices for Absorber Columns	100
5-7   Prices for Adsorber Platforms and Ladders	100
5-8   Total Capital Costs for Cold Water Condenser Systems	101
5-9   Additional Capital Cost for Refrigerant Condenser Systems	101
5-10  Prices for Negative Pressure, Insulated Fabric Filter Systems,
      Less Bags	101
5-11  Prices for Insulated Electrostatic Precipitators	102
5-12  Prices for Venturi Scrubbers	102
5-13  Required Steel Thicknesses for Venturi Scrubbers	102
5-14  Price Adjustment Factors for Venturi Scrubbers	102
5-15  Carbon Steel Straight Duct Fabrication Price
      at Various Thicknesses	103
5-16  Stainless Steel Straight Duct Fabrication Price
      at Various Thicknesses	103
5-17  Fan Prices	104
5-18  Carbon Steel Stack Fabrication Price for W Plate	105
5-19  Carbon Steel Stack Fabrication Price for 5/16" and %" Plate	105
                                  IX

-------
                                        Nomenclature*

a        =  packing constant
A        =  heat exchanger surface area, ft2
Abed     =  carbon bed cross sectional area, ft2
Acoiumn  =  absorber column cross sectional area, ft2
ACOn     =  condenser surface area, ft2

Anc      =  net cloth area, ft2
Ap       =  collection plate area, ft2
A,       =  venturi scrubber throat area, ft2
Atc       =  total cloth area, ft2
ABS     =  abscissa (Figure 4.7-2)

AC       =  adsorption capacity of carbon bed, Ib HAP/100 Ib carbon
A/C      =  air to cloth ratio for  baghouse, acfm/ft2
AF       =  absorption factor
b        =  packing constant
c        =  packing constant

C        =  annual credits, $/yr
Creq      =  amount of carbon required, Ib
Cpair     =  average specific heat of air, Btu/scf-°F
Cpair     =  average specific heat of air, Btu/lb-mole-°F
         =  average specific heat of combined gas stream, Btu/scf-°F

 .	 =  average specific heat of coolant, Btu/lb-°F
Cpe      =  average specific heat of emission stream, Btu/scf-°F
Cpe      =  average specific heat of emission stream, Btu/ lb-°F
Cpf      =  average specific heat of supplementary fuel (natural gas), Btu/lb-°F
Cptg     =  average specific heat of flue gas, Btu/scf-°F
Cpfig     =  average specific heat of flare gas, Btu/lb-°F
Cpw      =  average specific heat of water, Btu/lb-°F
CPHAP    =  average specific heat of HAP, Btu/lb-mole-°F
CE       =  collection efficiency (based on mass), %
CRF      =  capital recovery factor

CRFW    =  weighted average capital recovery factor
d        =  packing constant
D        =  annual direct labor costs, $/yr
Dbed     =  carbon bed diameter, ft
Dcoiumn  =  absorber column diameter, ft
Dduct     =  duct diameter, in
Dp       =  mean particle diameter, /urn
D,       =  venturi scrubber throat diameter, ft
Dtip      =  flare tip diameter, in
DG       =  diffusivity in gas stream, ft2/hr
aEnglish units are used throughout this report. Many engineering handbooks provide
 conversion factors for English to metric units.

-------
DL
Di
D2
DE
L't reported

DP
ex
f
FE
PER

9
9c
G
^area
Garea.f


Gmol
hd
he
hf
hflg

AH
'I con
Hf
H|0ad
rlnoncon

"uncon
HG
HL
HOG
Htcolumn

Httotal
HAPcon
HAPe
HAPe,m
HAP0

HAP0.m
HP
HR
HRS
L

L"
Lgal
U/Qe,a

LEL
m
M
Me
diffusivity in liquid, ft2/hr
annual operating labor cost, $/yr
annual supervision labor cost, $/yr
destruction efficiency, %
reported destruction efficiency, %
stream dew point, °F
excess air, % (volume)
fraction
fabricated equipment cost index
fan electricity requirement, kWh
packing constant
gravitational constant, = 32.2 ft/sec2
gas (emission stream) flow rate, Ib/hr
gas (emission stream) flow rate based on column cross sectional area, Ib/sec-ft2
gas (emission stream) flow rate at flooding conditions based on column cross sectional
area, Ib/sec-ft2

gas (emission stream) flow rate, Ib-mole/hr
heat content of emission stream after dilution, Btu/scf
heat content of emission stream, Btu/scf
lower heating value of supplementary fuel (natural gas), Btu/scf
flare gas heat content, Btu/scf

heat of vaporization of HAP, Btu/lb-mole
enthalpy change associated with condensed HAP, Btu/min
supplementary heat requirement (heat supplied by the supplementary fuel), Btu/min
condenser heat load,  Btu/hr
enthalpy change associated with noncondensible vapors, Btu/min

enthalpy change associated with uncondensed HAP, Btu/min
height of a gas transfer unit, ft
height of a liquid transfer unit, ft
height of a gas transfer unit (based on overall gas film coefficients), ft
absorber column packed height, ft

absorber column total height, ft
quantity of HAP condensed, Ib-mole/min
inlet HAP concentration, ppmv
quantity of HAP in the emission stream entering the condenser, Ib-mole/min
outlet HAP concentration, ppmv
quantity of HAP in the emission stream exiting the condenser, Ib-mole/min
fan power requirement, hp (horsepower)
heat recovery in the heat exchanger, %
number of hours of operation per year
solvent flow rate, Ib/hr
solvent flow rate based on absorber column cross sectional area, Ib/hr-ft2
solvent flow rate, gal/min
solvent flow rate, Ib-mole/hr
liquid flow rate in venturi scrubber, gal/min
liquid to gas ratio, gal/103 acf
lower explosive limit, % (volume)
slope of the equilibrium curve
annual maintenance costs, $/yr
moisture content of emission stream, %(volume)
                                               XI

-------
M2
M3
MWavg
MWe

MWflg
MWso|Vent
MWHAP

N
NOG

02
ORD
AP
APa
Pe

Ppartial
"vapor
AP.otal
APV
PC

Qa
Qc
'-'co
   olant
Qe,a
Qe,s
Of
Qfg
Qfg,a

Qfig
Qflg,a
Qrec
Qs
Qw

r
R

""•hum
Ref
RE
S
S
ScG
ScL

St
SV
annual maintenance labor cost, $/yr
annual maintenance supervision cost, $/yr
annual maintenance materials cost, $/yr
average molecular weight of a mixture of components, Ib/lb-mole
average molecular weight of emission stream, Ib/lb-mole

average molecular weight of flare gas, Ib/lb-mole
molecular weight of solvent, Ib/lb-mole
molecular weight of HAP (average molecular weight if a mixture of HAPs is present),
Ib/lb-mole
number of carbon beds
number of gas transfer units (based on  overall gas film coefficients )

oxygen content of emission stream, %  (volume)
ordinate (Figure 4.7-2)
total pressure drop for the control system, in H20
absorber column pressure drop, Ib/ft2-ft
emission stream pressure, mm Hg

partial pressure of HAP in emission stream, mm Hg
vapor pressure of HAP in emission stream, mm Hg
absorber column total pressure drop, in H20
pressure drop across venturi, in H20
purchased equipment cost, $

flow rate of gas stream at actual conditions, acfm
combustion air flow rate, scfm
flow rate of combined gas stream entering the catalyst bed, scfm
coolant flow rate, Ib/hr
cooling water flow rate, Ib/min

emission stream flow rate at actual conditions, acfm
saturated emission stream flow rate, acfm
supplementary fuel ( natural gas ) flow  rate, scfm
flue gas flow rate, scfm
flue gas flow rate at actual conditions, acfm

flare gas flow rate, scfm
flare gas flow rate at actual conditions,  acfm
quantity of HAP recovered, Ib/hr
steam flow rate,  Ib/min
cooling water flow rate, gal/min
packing constant
gas constant, = 0.73 ft3-atm/lb-mole °R;
relative humidity, %
refrigeration capacity, tons
removal efficiency, %

reported removal efficiency, %
packing constant
annual cost of operating supplies, $/yr
Schmidt number for HAP/emission stream
Schmidt number for HAP/solvent system

steam ratio, Ib steam/lb carbon
space velocity, hr"1
cleaning interval, min
=  1.987cal/g-mole°K

-------
tr
T
To
Tci
 1 con
 ' cool.i
 'cool.o
T6
 'fig
The
Tr

Tsti
Tsto
 1 wo
ATLM
 ' ncolumn
U
Uduct
ue
 ••e.s
  fig
U

'-'max
Ut
vc
"carbon
Vbed
"packing
w
*Vtco|urrln
x
X
y
Y
Y

Zbed
e
X
•n
Pbed
PC
PG
PL
9ads
Greg
"dry-cool
M-L
 residence time, sec
 temperature, °F
 combustion temperature, °F
 temperature of combined gas stream entering the catalyst bed, °F
 temperature of flue gas leaving the catalyst bed, °F

condensation temperature, °F
 inlet temperature of coolant, °F
 outlet temperature of coolant, °F
 emission stream temperature, °F
 temperature of saturated emission stream, °F

 flare gas temperature, °F
 emission stream temperature after heat exchanger, °F
 reference temperature,  = 70 °F
 inlet steam temperature, °F
 condensed steam outlet temperature, °F

 inlet cooling water temperature, °F
 outlet cooling water temperature, °F
 logarithmic mean temperature difference, °F
 absorber column thickness, ft
 overall heat transfer coefficient, Btu/hr-ft2- °F

 drift velocity of particles, ft/sec
 velocity of gas stream in the duct, ft/min
 emission stream velocity through carbon bed, ft/min
 throat velocity of saturated emission stream, ft/sec
 flare gas exit velocity, ft/sec
 maximum flare gas velocity, ft/sec
 annual utility costs, $/yr
 combustion chamber volume, ft3
 volume of carbon bed, ft3
 catalyst bed requirement, ft3
 absorber column packing volume, ft3
 particle grain loading, gr/acf
 absorber column weight, Ib
 mole fraction of solute in solvent, moles solute/(moles solute +  moles solvent)
 mole fraction of gaseous component in liquid, moles  solute/ moles solvent
 mole fraction of solute in air, moles solute/(moles solute + moles air)
 packing constant
 mole fraction of solute in air, moles solute/moles air
carbon bed depth, ft
packing constant
latent heat of vaporization for steam, Btu/lb
fan efficiency,  percent
density of carbon bed, Ib/ft3
density of carbon steel plate, Ib/ft3
density of gas  (emission stream), Ib/ft3
density of solvent,  Ib/ft3
cycle time for adsorption, hr
cycle time for regeneration, hr
cycle time for drying and cooling the bed, hr
viscosity of solvent, centipoise
viscosity of solvent, Ib/ft-hr

-------
                                           Chapter 1
                                         Introduction
1.1  Objective
The  objective of this handbook is to present  a
methodology for determining the performance and
cost of air pollution control techniques for reducing
or eliminating the emission of potentially hazard-
ous air pollutants (HAP's) from industrial/commer-
cial sources. (Note: The term "hazardous" in this
document is  very broad.  It is not limited to the
specific compounds listed under current regula-
tions [i.e., the Clean Air Act, the Resource Conser-
vation and Recovery Act, and the Toxic Substances
Control Act].) This handbook is to be used by EPA
regional, State, and local air pollution control agen-
cy technical personnel for two basic purposes: (1)
to respond to inquiries from interested parties (e.g.,
prospective permit applicants) regarding the HAP
control requirements that would be needed at  a
specified process or facility,  and  (2) to evaluate/
review permit applications for sources with the po-
tential  to emit HAP's. It should  be  noted that this
document provides general technical guidance on
controls and does not provide  guidance for compli-
ance with specific regulatory requirements for haz-
ardous air pollutants. Specifically, it does not speci-
fy design requirements  necessary  to achieve
compliance with standards established under spe-
cific programs such as Section 112 of the Clean Air
Act or standards established  under the Resource
Conservation  and  Recovery  Act.  Such require-
ments  vary with the hazardous  air pollutant  emit-
ted and with the emission source; thus, regulatory-
specific detailed specifications are beyond the
scope of this handbook.

Section 1.2 discusses the use  of this handbook.
Chapter 2 assists the user in identifying HAP's and
their respective emission sources.  Chapter 2 also
identifies the  key emission stream characteristics
necessary to select appropriate control techniques.
Chapter 3 provides additional  information to assist
the user in the control technique selection process
for each  HAP  emission source/stream. Chapter  4
presents  simple  step-by-step procedures to  deter-
mine basic design parameters of the specific con-
trol  devices and auxiliary equipment. Chapter  5
provides the necessary data and procedures to de-
termine order-of-magnitude  estimates (-60 to
+ 30 percent) for the capital and annualized costs of
each control system. Appendices A and B present
supplementary data and calculation procedures.
Appendix C contains blank worksheets to be used
while performing the functions described in this
handbook. These  worksheets are masters  from
which to make copies.

Additional appendices can be found in Evaluation
of Control Technologies for Hazardous Air Pollu-
tants—Appendices (EPA 600/7-86-009b; NTIS order
no. PB  86-167/038/AS; $22.95,  price subject to
change). The additional appendices further clarify
and expand the text and give derivations of equa-
tions, calculation procedures, and unit conversion
techniques. They are referenced in this handbook.
The  volume can  be ordered  from the  National
Technical Information  Service (NTIS),  5285 Port
Royal Road, Springfield, Virginia 22161;  (703)
487-4650.

A good  source of current information pertaining to
HAP's  is the "Air Toxics Information  Clearing-
house," which was established by EPA in response
to State and local agency requests for assistance in
the exchange of information on toxic air pollutants.
The Clearinghouse  is operated by EPA's Office of
Air Quality  Planning and Standards (OAQPS) in
close coordination with the State and Territorial Air
Pollution  Program  Administrators (STAPPA) and
the Association of  Local Air Pollution Control Offi-
cials (ALAPCO). The Clearinghouse collects and
disseminates information  from  State and  local
agencies, as well  as making users aware of air
toxics information available from  EPA and other
Federal  agencies. Specifically, the Clearinghouse
collects the following  air toxic information  from
State and local agencies: regulatory program de-
scriptions, acceptable ambient concentrations on
ambient standards, toxic pollutant research, source
permitting,  ambient monitoring,  toxicity testing,
and source testing.

The Clearinghouse  provides an on-line data base
containing all toxic-related information submitted
by State and local agencies, bibliographic citations
for relevant reports by EPA and other Federal agen-
cies, and references for ongoing EPA air toxic pro-
jects. It  also publishes a quarterly newsletter with
articles  on current  air toxics concerns. Finally, the
Clearinghouse periodically publishes various spe-
cial reports on topics of interest to users. For fur-

-------
ther information regarding the "Air Toxics Informa-
tion Clearinghouse," contact the appropriate EPA
regional office air toxics  contact,  or EPA/OAQPS,
Pollutant Assessment Branch, MD-12, Research Tri-
angle Park, North Carolina 27711; (919) 541-5645 or
FTS 629-5645.

1.2 How to Use the Handbook
Figure 1-1 is a flowchart of the steps performed
when responding to inquiries; Figure 1-2 presents
the same  type of flowchart when reviewing per-
mits. As shown by these figures, these two func-
tions are basically the same; the only substantive
difference is that the review process also compares
the determined/calculated parameters with the cor-
responding parameters stated in the permit appli-
cation to  ensure that the control system(s) pro-
posed by  the applicant will  provide the required
reduction of HAP emissions.

Once  an inquiry or permit application is received,
determine the HAP's applicable to the source cate-
gory in question  (Section 2.2). The HAP's are cate-
gorized under four headings: organic vapor, organ-
ic particulate,  inorganic vapor, and  inorganic
particulate. (Note:  For each HAP  group, a  list of
potentially or suspected  hazardous compounds
that may  be emitted  as  a HAP from  the  source
category is provided. This listing is neither all-inclu-
sive nor a  declaration that the compounds present-
ed  are hazardous.) Next, identify  the  potential
emission sources for each HAP group (Section 2.2).
The HAP emission sources are  listed under one of
three  classifications: process point  sources, pro-
cess fugitive sources, and area fugitive sources.
(Note: See  Section  2.2  for classification  defini-
tions.) After each emission source is determined,
identify the key HAP emission stream characteris-
tics (e.g.,  HAP concentration,  temperature, flow
rate, heat content, particle size) needed to select
the appropriate control technique(s) (Section 3.2).
Obtain the actual values for these characteristics
from  the owner/operator or from available litera-
ture if the owner/operator cannot provide the nec-
essary data. If two or more emission streams are
combined prior to entry into an air pollution control
system, determine the characteristics of the com-
bined emission stream ( Appendix B.1).

Depending upon the  specific  regulation and the
type/characteristics  of the HAP emission  source/
stream, the remaining steps in the methodology
will differ. There are four basic  "formats" for a
regulation: (1) a particular "control device" may be
required, (2) a "numerical limit" may be specified,
(3) a "technology forcing" requirement may be im-
posed, and  (4) a specific work practice or "other"
related practice  may  be  required. The regulation
format will define the steps that lead to the selec-
tion of the  appropriate  control technique(s). The
"control device" and "other" formats specify the
appropriate control technique(s). A "numerical lim-
it" format requires the determination of the HAP
removal efficiency before the appropriate control
technique(s) can be identified. Lastly, the "technol-
ogy forcing" format has two paths: one where the
cost of the control system is a factor in the decision,
and one where cost is not a factor. If control system
cost is a factor, the agency must determine the cost
constraints that will be imposed on the control
technique selection process. The steps that occur in
defining the HAP control requirements will depend
upon  each agency's regulatory policies.

The HAP emission stream characteristics, in con-
junction with the limitations imposed by the appli-
cable  regulations, are used to select the appropri-
ate control techniques (Chapter 3) for each HAP
emission  source/stream.  General  guidelines are
provided that match specific control  devices with
specified emission  stream properties  (e.g., HAP
content, temperature, moisture, heat content, parti-
cle size, flow rate). Basic  design  parameters are
then determined to  provide general design condi-
tions  that should  be met or exceeded for each se-
lected control  technique to achieve the specified
HAP removal efficiency (Chapter 4). This exercise
also identifies  which of the selected  control tech-
niques will  not achieve the desired  HAP control
requirements.  The basic  design parameters also
can be used to obtain an order-of-magnitude cost
estimate for each control  device  (Chapter  5). As
noted above, this cost information can be an inte-
gral part of the HAP control system selection pro-
cess.  After completing the above process, a HAP
control program can be recommended or evaluat-
ed. (For an example of a State HAP program, see
Appendix A.1, reference 1.)
                 Example Case
 To guide the user through the steps and calcula-
 tions described in this handbook, examples are
 provided throughout the text. As shown here,
 each example is boxed. The primary example
 case pertains to a hypothetical  plant owner re-
 questing assistance  in determining  the type of
 control system that should be used on an emis-
 sion stream generated by a paper coating drying
 oven. This example is carried through the entire
 handbook. Additional example  emission
 streams are introduced in Chapter 3 to illustrate
 fully the control technique selection process and
 to clarify the design procedures of Chapter 4.
1.3  References
1. U.S. EPA. Evaluation  of Control Technologies
   for  Hazardous Air Pollutants  — Appendices.
   EPA-600/7-86-009b (NTIS PB 86-167/038/AS). Oc-
   tober 1985.

-------
Figure 1-1.   Steps used when responding to inquiries.
                Manual Location to
                  Perform Step
                  1Section 2.2
                  2Section 2.3
                  3None
                  ""Appendix B.1
                  5None
                  6Chapter 3
                  7Chapter 4
                  8Chapter 5
                                                Information Requested
                                                 on HAP Control for
                                                  a Specific Facility
                 Obtain Available
                Plant-Specific Data
                       I
                  Define HAP's1
                       _L
              Define Emission Sources
                Generating HAP's1
                       !
                                                Define Characteristics
                                                Needed for Each HAP
                                                  Emission Stream2
                                                        ±
              Combine HAP Streams3
                                                                M
                                                                             Inquirer Assistance
                                                Define Characteristics
                                                Of Combined Streams4
                                                        ±
                                               (Each Single/Combined Y^_
                                                HAP Emission Stream JT~

                                               ~     I
                                                 Define HAP Control
                                                   Requirements6
        Control
        Device
Numerical
  Limit
       Select Appropriate
     Control Techniques(s)6
              ±
  Determine Required
  Control Efficiency5
         Are
Basic Design Parameters
      Requested       /
Determine Basic Design
    Parameters for
  Control System(s)7
                                                                   Technology
                                                                     Forcing
                                                           Cost of Control a
                                                            Decision Factor
             Other
          Requirement
Regulation Agency
  Policy Decision
                                               Agency Determination
                                                of Cost Constraints5
                                           Select Appropriate
                                          Control Technique(s)6
                                                 Select Appropriate
                                               Control Technique(s)6
                                                        ±
                                               Determine Basic Design
                                               Parameters and Cost for
                                                Control System(s)7'8
                                                         t
                                               Select Control System
                                               Having Most Stringent
                                               Level of Control Within
                                               Given Cost Constraints
                                                        A.
                                              Recommend Appropriate
                                                Control Technique(s)
                                                        _L
                                                                                   J_
                                         Determine Basic Design
                                             Parameters for
                                           Control System(s)7
                                                  ±.
                                          Select Control System
                                          Having Most Stringent
                                             Level of Control
                                                        Last
                                                    HAP Emission
                                                      Stream
                                     No
\
Yes
Recommend HAP
Control Program

-------
Figure 1-2.   Steps used when reviewing permits.
          Manual Location to
             Perform Step
            1Section 2.2
            2Section 2.3
            3None
            "Appendix B.1
            6None
            6Chapter3
            'Chapter 4
            8Chapter 5
Permit Application
for Review/Approval


/ Are All HAP's \
\ Addressed1 ? /
Yes


/ Are All HAP Emission \
\ Sources Addressed1 ? /
Yes
i

i
No

No


Obtain Additional Data
from Applicant



Obtain Additional Data
from Applicant



                                                 Are All HAP Emission
                                               Stream Characteristics
                                                   Provided2 ?
                                        No
                                                      Obtain Additional Data
                                                         from Applicant
                                                    Yes
                                                    Are Any HAP
                                                 Emission Streams
                                                   Combined3 ?
                                       LYes
                          Are Combined Stream
                            Characteristics
                              Correct4 ?
                                                      No
                                                                                              Yes
                                                Each Single/Combined
                                                HAP Emission Stream
1

Define HAP Control
Requirements5
                                                      Obtain Additional Data
                                                          from Applicant
Control
Device ,

Select Appropriate
Control Techniques(s)6

Numerical
Limit ,

Determine Required
Control Efficiency5


Other
Requirement
Regulatory Agency
Policy Decision5
Technology
Forcing
Yes/ Is Cost of Control a \
r\ Decision Factor ? /
   Determine Basic Design
       Parameters for
      Control System(s)7
   /         Is
    Permit Control System
        Appropriate
Recommend Appropriate
  Control Technique(s)
               Yes
    Permit System Design
         Appropriate
Recommend Appropriate
Basic Design Parameters
               Yes
 Agency Determination
  of Cost Constraints5
1
No
Select Appropriate
Control Technique(s)6
                               Select Appropriate
                             Control Technique(s)6
Determine Basic Design
Parameters and Cost for
  Control System(s)7'8
                            Determine Basic Design
                                Parameters for
                              Control System(s)7
                                                             Select Control System
                                                             Having Most Stringent
                                                             Level of Control Within
                                                             Given Cost Constraints
                                                                     J_
                                                          Select Control System
                                                          Having Most Stringent
                                                             Level of Control
                                                 Last HAP Emission
                                                     Stream ?
                                        No
1
Yes
Permit Approval or
Provide Recommendations

-------
                                           Chapter 2
                     HAP Emissions and Their Key Physical Properties
2.1  Background
This chapter's primary goal is to identify the follow-
ing: (1) potential HAP'sfor a given source category
and the specific sources that may emit the potential
HAP's—Section 2.2, and (2) key emission stream
physical  properties needed to select  appropriate
control strategies and size control devices for the
HAP emission sources Section 2.3. Specific source
categories are divided into nine general classifica-
tions in this manual. (Note:  The general classifica-
tion system is a hybrid of the classification  systems
used by  the  following documents:  (1) U.S.  EPA,
Compilation of Air Pollutant Emission Sources —
Third Edition: Supplements 1-15, AP-42,  January
1984; and (2) the U.S. EPA, BACT/LAER Clearing-
house—A Compilation of Control Technology De-
terminations, April  1983.) Every possible source
category cannot be listed; however, similarities ex-
ist between many categories. Thus, the user should
be able to obtain some  guidance for any specific
facility. Only those source categories that are
known to emit potential HAP's are presented in this
handbook. (For a listing of chemical hazard infor-
mation profiles  [CHIP's] and CAS numbers, see ap-
pendix A.2, reference 1.)

Individual source categories have been classified
based on the manufacturing process associated
with emissions of potential HAP's. The Solvent Us-
age  Operations  classification includes processes
dependent  on solvents, such  as surface  coating
and  dry cleaning operations.  Metallurgical Indus-
tries include processes associated with the manu-
facture of metals, such as primary aluminum pro-
duction. Processes and operations associated with
the manufacture of organic and inorganic chemi-
cals have been grouped into the Synthetic  Organic
and  Inorganic Chemical Manufacturing classifica-
tions, respectively.  Industries using chemicals in
the formulation of products are  classified as
Chemical Products Industries.  The Mineral and
Wood Products  Industries classifications  include
operations such as asphalt batch plants and kraft
pulp mills, respectively. The Petroleum Related In-
dustries classification is defined as oil and gas pro-
duction, petroleum refining, and basic petrochemi-
cals  production. Combustion  Sources are utility,
industrial, and residential combustion sources us-
ing coal, oil, gas, wood, or waste-derived fuels.
To assist the user in recording the pertinent infor-
mation, a worksheet has been provided. A copy of
this  worksheet, the HAP  Emission Stream  Data
Form, is presented in Appendix C.1. An example of
a partially completed worksheet is shown in Figure
2-1. This worksheet is designed to record informa-
tion pertaining to one emission stream, be it a sin-
gle stream or a combined stream consisting of sev-
eral single streams.
                 Example Case
  Information has been requested by a paper coat-
  ing  plant owner regarding the control  of an
  emission stream from his facility's drying oper-
  ations. The  most likely generic classification to
  include a paper coating plant (one of many sur-
  face coating industries) would be Solvent Usage
  Operations. To determine if a paper coating fa-
  cility is listed  within this category, see Section
  2.2.1  and Table 2-1, which indicates that the ini-
  tial choice was correct (i.e.,  paper coating  is list-
  ed under SC-Paper, Tapes, Labels) and the infor-
  mation retrieval process can begin.
2.2  Identification of Potential HAP's and
     Emission Sources
The  purpose of this  section is to  present general
information  on emissions  of potential HAP's  by
source category. Within each of the nine general
classifications, information  is presented on  the
types of potential HAP's that may  be emitted by a
particular source category. This  information  in-
cludes the names of specific compounds, the clas-
sification of the compounds (i.e.,  organic or inor-
ganic), and  the form in which these compounds
would be emitted (i.e., vapor or particulate). Owing
to process variations, actual emissions from specif-
ic facilities may differ from the general information
presented.  Complete  identification of HAP emis-
sions is best accomplished with  assistance from
the owner/operator of the facility. (See Appendix
A.3,  reference 1, for  a listing of trade  names and
common synonyms for HAP's.)

This section also presents  information  pertaining
to the sources (e.g., processes) within each specific
source category that have the potential  to  emit

-------
Figure 2-1.  A partially completed HAP Emission Stream Data Form for one of six emission streams (#1) generated at a fictitious
          company.
HAP EMISSION STREAM DATA FORM*
Cor
Loc
A.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
O.
npany Glaze Chemical Company
atjnn ( nf Fmissinn Streams Under Review

ant Identification *1 /
(a) paper coating oven
(a) process point
,a) toluene
(a) organic vapor
(a) 73 ppmv
(a) 28.4 mm Hg (3 77°F
(a) insoluble in water
|a\ provided
(a) 92 Ib/lb-mole
2% vol.
12C°F
15,000 scfm (max)
atmospheric
none / none

#3 Oven Exhaust
(b)

(b)
(b)
(h)
(h)

(h)
(b»
P. Organic Content (1)***
Q. Heat/02 Content (1)
R. Paniculate Content (3)
S Particle Mean Diam (3)
T. DriftVelocitv/SO-,(3)


, >
(r)
(c)
(c)

(r)
(n)
(r)
(r.\
44 ppmv CH/j, 4 pomv ot
0.4 Btu/scf / 20.6% vol.


/
 U.  Applicable Regulation(s)	
 V.  Required Control Level	
 W.  Selected Control Methods
   *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
    method(s). Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

  **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 = organic vapor process emission
      2 = inorganic vapor process emission
      3 = participate process emission

 l»* Organic emission stream combustibles less HAP combustibles shown on Lines D and F.
 HAP's. In this  handbook,  emission sources  are
 broadly classified into three groups: process point
 sources, process fugitive sources, and area fugitive
 sources. Point sources are, in general, individually
 defined.  Reactors, distillation columns, condens-
 ers, furnaces, and boilers are typical point sources
 which discharge emissions through a vent-pipe or
 stack. These sources can be controlled through the
 use of add-on  control  devices. Process  fugitive
 sources, like process point sources, are individually
 defined.  Emissions from  these sources  include
 dust, fumes, or gases that escape from or  through
 access ports and feed or discharge openings to a
 process (e.g., the open top of a vapor degreaser).
 Process fugitive sources also can include vent fans
 from rooms or enclosures containing an emissions
 source (e.g., a  vent fan on a dry  cleaner). These
 sources can be controlled by add-on control de-
 vices once the emissions are captured by hooding,
 enclosures, or closed vent systems and then trans-
 ferred to  a control device.

 Area  fugitive sources are  characterized by large

                        6
surface areas from which emissions occur. In addi-
tion, process  equipment  such as pumps, valves,
and compressors  are considered  area fugitive
sources;  emissions  from  these sources  occur
through leaks during  process operation. Although
these sources are small, they are usually found in
large numbers dispersed over a wide  area  in a
given process. Area fugitive sources also include
large and undefined  emission sources such as
waste  treatment lagoons,  raw  material storage
piles, roads, etc.
                                       •
The sources listed in this handbook generate emis-
sions; however, a definitive statement as to wheth-
er they emit a HAP  cannot be made. As in the case
of identifying the potential HAP's emitted at a spe-
cific facility,  communication  with the owner/
operator is useful in  identifying each  source that
emits a HAP. The listings  found in this section are
not all-inclusive; a specific  facility  may have an
emission  producing operation that is not common
to its industry and, thus, the source may not be
included here.

-------
Table 2-1.    Potential HAP's and Emission Sources for Solvent Usage Operations

                                          Potential HAP's8
Source Key
A — bath evaporation
B —solvent transfer
C —ventilation
D —waste solvent disposal
E —washer
F —drying
G —still, filtration
H — cooker
                                       Potential Emission Sources
Source Category
Solvent Degreasing
Dry Cleaning
Graphic Artsb
Waste Solvent Reclaiming
SCc-Flatwood Panelingd
SC-Machinerye
SC-Appliances'
SC-Metal Furniture
SC-Auto/Truck9
SC-Fabrics
SC-Cansh
SC-Paper, Tapes, Labels
Magnetic Tape Coating
SC-Electrical Insulation
SC-Marine Vessels'
Vinyl & Acrylic Coatings'
SC-Wood Furniture
SC-Trans. Vehicles"
Machine Lubricants
Rubber Tire Manufacturing
Organic Inorganic
Vapor Particulate Vapor Particulate
X
X
X
X
X
X X
X X
X X
X X
X
X
X
X
X
X X
X
X
X X
X
X
Process
Point
C
F
F
N
0
0
0
0
0
0
0
0
F
0

0
0


F
Process
Fugitive
A,B,D
E,G,H,I
L,M,N
I
UP
Q,R
Q,R
Q,R
S,R
D,K,Q,R,T
Q,U
B,I,Q,T
I,Q,T
Q

UP
S,R,T
S,R

I,V,W
Area
Fugitive
K
J,K

J,K










Q



Q

I  — solvent storage
J — pipes, flanges, pumps
K —transfer areas
L — rollers
M— ink fountains
N — condenser
O — oven
P —coaters
Q — application area
R —flashoff area
S — spray booth
T — solvent/coating mixing
U —quench area
V — green tire spraying
W— sidewall/tread end/undertread
    cementing
"References 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,13, 14, 15,18, 31, 32.
blncludesflexography, lithography, offset printing, and textile printing.
CSC: surface coating.
dlncludes coating of other flat stock.
"Includes coating of misc. metal parts, machinery, and equipment.
' Includes all categories of appliances, large and small.
9lncludes coating of automobiles and light-duty trucks.
hlncludes surface coating of coils, cans, containers, and closures.
' Includes coating and maintenance of marine vessels.
' Includes vinyl, acrylic, and nitrocellulose coatings.
"Includes coating of trucks, buses, railroad cars, airplanes, etc.
2.2.1  Solvent Usage Operations
Solvent usage operations are defined as manufac-
turing processes that use solvents, including such
processes as surface coating operations, dry clean-
ing, solvent degreasing, waste solvent reclaiming,
and graphic arts. Table 2-1 lists source categories
within this  group  of operations that have been
identified as sources of volatile organic compound
emissions that may include potential HAP's. As is
shown by Table 2-1, all solvent usage operations
generate organic vapor emissions. (Note: Some of
the emission sources  generate aerosols [i.e., or-
ganic particulate]; however, the aerosols evaporate
in a short time  and the emissions normally are
controlled as a vapor. Therefore, Table 2-1 does not
indicate the presence of organic particulates.) Due
to the large number of potential HAP's associated
with these types of operations, the format of Table
                   2-1 prohibits the inclusion of compound-specific
                   data. Potential  HAP's  that may be emitted  by
                   sources in Table 2-1 are summarized in Appendix
                   A.1; this appendix  lists both specific compounds
                   and classes of compounds that may be emitted by
                   sources within the category. Appendix A.1 can be
                   used to determine whether a particular solvent us-
                   age operation may emit a specific potential HAP or
                   group of potential HAP's, as well as to determine all
                   solvent use operations that may emit a particular
                   potential HAP. Table 2-1  presents the  emission
                   sources that may emit potential HAP's.

                   2.2.2  Metallurgical Industries
                   The metallurgical industries can be broadly divided
                   into primary, secondary, and miscellaneous metal
                   production operations. The majority of this indus-
                   try is covered  under SIC Codes 331, 332, 333, 334

-------
                                         Example Case
  As directed  by Section 2.2.1, Appendix A.1 is
  used to determine the potential HAP's. The po-
  tential HAP's for paper coating operations are as
  follows:
  Specific Compounds       Generic Compounds
  toluene                  mineral spirits
  xylene                   other aromatics
  ethylene glycol           alcohols
  acetone                  cellusolves
  methyl ethyl ketone       ketones
  methyl isobutyl ketone    esters
  ethyl acetate
  Upon reviewing data  from the solvent vendor,
  the owner determined that only toluene is pres-
ent in the solvent being evaporated  by the
ovens. Table 2-1 indicates that toluene is an or-
ganic compound, and it would be emitted as a
vapor. This information is then listed on the HAP
Emission Stream Data Form provided in  Appen-
dix C.I (see Figure 2-1).

Also for the example case, the source (oven: a
process point source) was identified. If the user
were interested in the other emission-releasing
processes at a paper coating operation, Table 2-
1 indicates that the remaining sources  include
solvent transfer,  solvent storage, application
areas, and solvent/coating mixing.
and 336. The term primary metals refers to produc-
tion of the metal from ore. The secondary metals
industry includes the recovery of metal from scrap
and salvage and the  production of  alloys  from
ingots. The miscellaneous subdivision includes in-
dustries with operations that produce or use metals
for final products. Table 2-2 presents the potential
HAP's for these industries and the industry-specific
emission sources.
2.2.3 Synthetic Organic Chemical
      Manufacturing Industry (SOCMI)
The SOCMI is a large and diverse industry produc-
ing several thousand intermediate and end-product
chemicals from a small number of basic chemicals.
Most of the chemicals  produced by this industry
fall under SIC Code 286. Due to the complexity of
the SOCMI, a general approach is used in this sec-
tion to describe generic emission sources and spe-
cific emission source types. This approach is identi-
cal to the approach used by EPA in its efforts to
develop new source performance standards for the
SOCMI.

A  large  proportion  of the emissions from the
SOCMI occur as organic vapors. However, organic
particulate emissions may be generated in some
processes  (usually  during the  manufacture of
chemicals that exist  as solids at  ambient  condi-
tions). The emissions typically contain raw materi-
als (including impurities) used in and intermediate
and final products formed  during the manufactur-
ing process. Many of these emission streams may
contain  HAP's due to the  great number of com-
pounds  manufactured in the SOCMI. (See Appen-
dix A.5, reference 1, for further information on spe-
cific emissions for different SOCMI processes.)
 Potential emissions from this industry can be de-
 scribed generically as follows:

 (a) Storage and handling emissions
 (b) Reactor process emissions
 (c) Separation  process emissions
 (d) Fugitive emissions
 (e) Secondary  emissions (e.g.,  from waste treat-
    ment).

 Emissions can potentially occur from raw materials
 and product storage tanks as working and breath-
 ing losses through vents. Emissions from handling
 result during transportation or transfer of the vola-
 tile organic liquids. Reactor processes and separa-
 tion processes are the two broad  types of pro-
 cesses  used in manufacturing organic chemicals.
 Reactor processes involve chemical reactions that
 alter the  molecular structure of chemical com-
 pounds. A  reactor process involves a reactor, or a
 reactor in combination with one or more product
 recovery devices. Product recovery devices include
 condensers, adsorbers, and absorbers.  Emissions
 from reactor processes occur predominantly from
 venting of inert gases from reactors and product
 recovery devices, or from  the release of organic
 compounds that cannot be recovered economical-
 ly. Typical  emission sources in reactor processes
 include point sources (e.g., vents on reactors and
 product recovery devices), process fugitive sources
 such as disposal of bottoms from the reactor or the
 product recovery devices, and area fugitive sources
 (e.g., pumps, valves, sampling lines, and compres-
 sors).

 Separation processes  often follow  reactor pro-
 cesses and divide chemical product mixtures into
 distinct fractions. Emissions from separation pro-
 cesses  are associated  primarily with absorption,
 scrubbing,  and  distillation operations. Other sepa-
                      8

-------
Table 2-2. Potential HAP's and Emission Sources for Metallurgical Industries

Source Category

Primary Aluminum Production

Primary Cadmium Production
Metallurgical Coke


Primary Copper Smelting

Ferroalloy Production

Iron and Steel Production

Primary Lead Smelting

Primary Zinc Smelting

Manganese Production
Nickel Production

Potential
Organic
Vapor Paniculate



3,8, 13, 19, 18
21,23,26,
27,28,30
18

18

18

18

18

18
18

Secondary Aluminum Operations
Secondary Copper Operations
(Brass and Bronze Production)
Gray Iron Foundries


Secondary Lead Smelting
Steel Foundries
Secondary Zinc Processing
Lead Acid Battery Production
18

2,3,13, 18
19,21,23

18



Cadmium-Nickel Battery Production
Dry Battery Production
Misc. Lead Products
Pollutant Key
7 — arsenic
2— acrolein
3— acetaldehyde
4 — ammonia
5 — antimony
6— barium
7 — beryllium
8 — benzene
9 — cadmium
70 — chromium
7 7 — copper
72 — fluoride
73 — formaldehyde
74— lead
75 — mercury




16 — manganese
17— nickel
18— polycyclic organic
matter (POM)
75 — phenol
20— selenium
27— toluene
22 — vanadium
23— xylene
24 — zinc
25 — iron
26— cresols
27— cyanides
28— pyridine
29 — hydrogen sulfide
30 — methyl mercaptan
HAP's0 Potential Emission Sources
Inorganic prorssK
Vapor Paniculate Point
12 72 A,I,J,
M,N,R
9 J,E
4,29 7,5,6,7,9 B
74,75,76
17,20,22
1, 12 1,5,9, 1 1 F,J,T
74,75,20,24
9, 10, 11,14 J
76,77,22,24
72 6,9,70,77 B,J,V
74,76,77,22,24
7, 12 1,5,9, 1 1 J,V
74,75,20
7,72 7,9,77 E,J,T,S
74, 75, 20,24
76 J
7,72 7,9,74 A,I,J,M,T
77,20,24
12 12, 17 J
24 9,77,74 J
77,20,24
7,6,7,9,70,77 J,Y
74,75,76,77
22,24,25
7,74,76,20 J
7,70,7,76,77,25 J,Y
24 9,75,77,20,24 J,E,S
74 74
9,74 V
76
74 5, 74
Source Key
A — calciner
B — coke oven
C — coke oven charging/
pushing
D — coke quenching
E —condenser
F — converter
G — converter charging/etc.
H — furnace tapping
J — furnace
K. — furnace charging
L — galvanizing vessel
M — material crusher/mill
N — material storage and
handling

Process Area
Fugitive Fugitive
H,K,D N,Q,U,Z

0,P N,Z
C,0,X N,D,Q,U


G,H,K,0,P,X N,Q,U,W,Z

H,K,0,P N,Q,W

C,H,K,0,X D,N,Q,U,W,Z

H,K,O,P N,Q,U,W,Z

0 N,Q,U,W,Z

H,K,M,P N,Q,Z
P N,Q.Z

H,K,P U
H,K,P U

H,K,G,P U


H,K,P U,Q
G,H,K,P U
H,K,L,P U
0,P
N,0
M,N,0
G,0,P

0 — material preparation
P — metal casting
Q — outdoor storage pile
R — reduction cell
S — retort
T — roaster
U — service road
V — sintering machine
W — slag dumping
X — vessel leakage
Y — foundry mold and core
decomposition
Z — mining operations



a References 6,16, 17, 19, 20, 21, 22, 23, 24, 25, 26, 28, 29, 30, 33, 34, 35.
                                                                                            9

-------
ration processes that may contribute to emissions
include drying, filtration, extraction, settling, crys-
talization, quenching, evaporation, ion exchange,
dilution, and  mixing/blending. One  of  the more
commonly employed separation techniques is dis-
tillation. Depending on the type of distillation sys-
tem used  (i.e.,  vacuum or  nonvacuum),  typical
emission points  can include condensers, accumu-
lators,  hot wells,  steam  jet ejectors, vacuum
pumps, and pressure relief valves. Emission points
from a  vacuum  distillation system  are  shown in
Figure 2-2.

Although fugitive emissions are listed as a separate
group, they can  occur from storage and handling,
reactor  processes, and separation processes. Area
fugitive sources include groups of valves, pressure
relief devices, pumps  and compressors, cooling
towers, open-ended lines, and sampling systems.
Process fugitive sources include hotwells, accumu-
lators, and process drains from  reactors, product
recovery devices, and separation equipment.

Table 2-3 presents  information on  specific emis-
sion points and emission source types for each of
Table 2-3.    Emission Sources for the SOCMI"
Generic Source Category
Potential Emission Sources (Specific)
  Process    Process     Area
   Point    Fugitive   Fugitive15
Storage and Handling
Reactor Processes
Separation Processes
Fugitives


E,F
F,L


A
G
G,M,N
G,M,N

B,C,D
C,D,H,I,J,K
K
B,C,D,H,I
J,K,M,N,0
Source Key
A — storage, transfer, and     I  —
    handling              J —
B—spills                 K —
C —valves                L —
D —flanges
E — reactors
F — product recovery devices
    (absorber, adsorber,     M —
    condenser)             N —
G — process drains          0 —
H — pumps
        compressors
        sampling lines
        pressure relief devices
        separation devices
        (distillation column,
        absorber, crystalizer,
        dryer, etc.)
        hotwell
        accumulator
        cooling tower
"References 12, 36, 37, 38, 39, 40, 41.
bGroups of small point sources (e.g., valves, compressors,
 pumps, etc.) at a SOCMI plant are considered as area fugitive
 sources in this handbook.
Figure 2-2.   Potential emission points (shaded) for a vacuum distillation column using steam jet ejectors with barometric
           condenser.(40)
                                                            Steam
                                                                                          Steam
                                                                                             I Ejector
                                                                     Hotwell
                        10

-------
the generic emission source groups. Using this in-
formation, the user can identify the potential emis-
sion  sources pertaining  to his specific situation.
(See Appendix A.5,  reference 1, for an illustration
of the approach outlined above and showing the
emission sources and the HAP's potentially emitted
from a SOCMI process.)

2.2.4 Inorganic Chemical Manufacturing
      Industry
This industry includes the manufacture of the basic
inorganic chemicals before they  are used in the
manufacture of other chemical products.  Most of
the chemicals produced  by this industry fall under
SIC Code 281. Potential emissions from these pro-
cesses may be high, but because of economic ne-
cessity they are usually recovered. In some cases,
the manufacturing operation is run as a closed sys-
tem, allowing little or no emissions to escape to the
atmosphere. Table 2-4 presents the potential HAP's
and industry-specific emission sources for  these
industries.

2.2.5 Chemical Products Industry
This industry includes the manufacture of chemical
products such as carbon  black,  synthetic fibers,
synthetic rubber and plastics, which may be used
in further manufacture. Also included are the man-
ufacture of finished chemical products for ultimate
consumption such as Pharmaceuticals, charcoal,
soaps and detergents; or products to be  used as
materials or supplies in  other industries such as
paints, pesticides, fertilizers and explosives. Most
of the chemical products are covered under  SIC
Codes 282,  283, 284, 285, 287 and 289. As  in other
chemical industries, the  potential emissions from
these processes may be  high, but because of eco-
nomic necessity they are usually recovered. Table
2-5 presents the potential HAP's and the industry-
specific emission sources for this industry.

2.2.6 Mineral Products Industry
This industry involves the processing and  produc-
tion of various nonmetallic minerals. The industry
includes cement production, coal cleaning and con-
version, glass and glass fiber manufacture,  lime
manufacture, phosphate  rock and taconite ore pro-
cessing, as well as various  other manufacturing
processes.  Most of this industry falls under  SIC
Codes 142,  144, 145, 147, 148, 149, 321, 322, 323,
324, 325, and 327. Table  2-6 presents the potential
HAP's and the industry-specific emission  sources
industries.

2.2.7 Wood Products Industry
The  wood  products industry  involves industrial
processes that convert  logs  to pulp, pulpboard,
hardboard,  plywood, particleboard,  or  related
wood products and wood preserving. This industry
falls under SIC Codes 242, 243,  249,  and 261.
Chemical wood pulping involves the extraction of
cellulose from wood by dissolving the lignin that
binds the cellulose fibers together. The principal
processes used in chemical  pulping are kraft, sul-
fite, and neutral  sulfite. Plywood  production  in-
volves manufacturing wood panels composed of
several thin wood veneers bonded together with an
adhesive. The wood preserving process is one in
which sawn wood products are treated by injection
of chemicals that have fungistatic and insecticidal
properties or impart fire resistance. Table 2-7 pre-
sents the potential HAP's and the industry specific
emission sources for these industries.

2.2.8 Petroleum Related Industries
In this handbook, the petroleum related industries
source category includes the oil and gas produc-
tion industry, the  petroleum  refining industry, and
the basic petrochemicals industry; these industries
fall under SIC Codes 13 and 29.

The oil and gas production  industry includes the
following processes: exploration and site prepara-
tion, drilling, crude processing,  natural gas pro-
cessing, and secondary or tertiary recovery. The
principal products of this industry are  natural gas
and crude oil.

The petroleum refining  industry involves various
processes that convert crude  oil into more than
2,500 products, including liquefied petroleum gas,
gasoline, kerosene, aviation fuel, diesel fuel, a vari-
ety of fuel oils, lubricating oils, and feedstocks for
the petrochemicals industry.  The different pro-
cesses involved in the petroleum refining industry
are crude separation, light hydrocarbon process-
ing, middle and heavy  distillate processing, and
residual hydrocarbon processing.

In the basic petrochemicals industry, hydrocarbon
streams from the oil and gas  production and petro-
leum refining  industries are converted into feed-
stocks for  the  organic chemical industry. These
feedstocks  include benzene,  butylenes, cresol and
cresylic  acids,  ethylene, naphthalene,  paraffins,
propylene,  toluene, and xylene.  The main pro-
cesses used by this industry  are separation, purifi-
cation, and chemical conversion processes. (See
Appendix A.6, reference 1, for a breakdown of typi-
cal processes involved  in each of the three indus-
tries.)

Table 2-8 presents the potential HAP's that may be
emitted  from these industries. Table 2-9 provides
more specific  information on potential emissions
from the petroleum refining industry  segment of
this generic category. A large  proportion of  the
emissions occur as organic  vapors; for example,
benzene, toluene, and xylenes are the principal or-
ganic vapor emissions. This is due to the chemical

                      77

-------
Table 2-4. Potential HAP's and Emission Sources for the Inorganic Chemical Manufacturing Industry
Potential HAP's Potential Emission Sources
Source Category
Aluminum Chloride
Aluminum Fluoride
Ammonia
Ammonium Acetate
Ammonium — nitrate, sulfate
thiocyanate, formate, tartrate
Ammonium Phosphate
Antimony Oxide
Arsenic — disulfide, iodide
pentafluoride, thioarsenate
tribromide, trichloride.
trifluoride, trioxide
orthoarsenic acid
Barium — carbonate, chloride
hydroxide, sulfate, sulfide
Beryllium — oxide, hydroxide
Boric Acid and Borax
Bromine
Cadmium (pigment) — sulfide
sulfoselenide, lithopone
Calcium— carbide, arsenate
phosphate
Chlorine
Chlorosulfonic Acid
Chromic Acid
Chromium — acetate, borides
halides etc.
Chromium (pigment) — oxide
Cobalt — acetate, carbonate
halides, etc.
Copper Sulfate
Fluorine
Hydrazine
Hydrochloric Acid
Hydrofluoric Acid
Iodine (crude)
Iron Chloride
Iron (pigment) — oxide
Lead — arsenate, halides
hydroxides, dioxide.
nitrate
Lead Chromate
Lead (pigments) — oxide
carbonate, sulfate
Manganese Dioxide
(potassium permanganate)
Manganese sulfate
Mercury — halides, nitrates,
oxides
Nickel — halides, nitrates,
oxides
Nickel Sulfate
Nitric Acid
Phosphoric Acid
Wet process
Thermal process
Phosphorus
Phosphorus Oxychloride
Inorganic
Vapor
4,W
17
1
1
1

1,17
5
2








8,10


3,17

10
19,24
12





14
17
1,39
19,20
17
10
10,20
40
3


22


24






27
28
10,17,2,30,18


17
10

Particulate








2




6

7
9

15

2

25

11,12
11

11
13




20

38
20

2,21



21

23

23
25

26

26
28
30




Process Process
Point Fugitive
X X
X X
B,D,E K
X X
C,F,I,L Q

X X
X X
H,U K,Q,T




C,E,G,I,L,U N,P,Q,T

X X
X X
X X
X X

H K,P

H,C K,R
X X
H K,N,0,Q
X X

X X
X X

X X
X X
X X
B
B,G K,R
X X
X X
X X
G,L P,Q


G,R P,Q
G,R P,Q

G,L Q,P,T

G,L Q,P,T
X X

P,Q

L Q,T
B,H K,N,R

H,C,W K,N,P,T
B,G K,N,R,T
X X
X X
Area
Fugitive


J,S





J,S













J

J,S



























J,S

J,S
J,S



-------
                                            Potential HAP's3
                                              Potential Emission Sources
 Source Category
        Inorganic
                                        Vapor
                                                          Paniculate
                                        Process
                                         Point
                                Process
                                Fugitive
                              Area
                            Fugitive
Phosphorus Pentasulfide
Phosphorus Trichloride
Potassium — bichromate,
   chromate
Potassium Hydroxide
Sodium Arsenate
Sodium Carbonate
Sodium Chlorate
Sodium Chromate —
   dichromate
Sodium Hydrosulfide
Sodium— silicofluoride,
   fluoride
Sulfuric Acid
Sulfur Monochloride —
   dichloride
Pollutant Key
 7— ammonia
 2— arsenic
 3— arsenic trioxide
 4—aluminum chloride
 5— antimony trioxide
 6— barium salts
 7— beryllium
 8— bromine
 9— boron salts
10— chlorine
77— chromium salts
12— chromic acid mist
13— cobalt metal fumes
14— copper sulfate
15—cadmium salts
76—chromates (chromium)
77—fluorine
18— hydrogen sulfide
79— hydrogen chloride
20— hydrochloric acid
21— lead
22— lead chromate
23— manganese salts
24— manganese dioxide
25— mercury
26— nickel
 29,31
32,10,29
   16

   10


    1
   10
   16

   18
   17

 33,34
   10
29
29
16

25
2
16
33
   X
   X
   I

   X
   H
  I,L,V
   X
G,I,L,M

   X
   X

A,B,C,H
   X
 X
 X
 X
K,P
 P
 X
P,Q

 X
 X

K,R
 X
J,S
Zinc Chloride
Zinc Chromate (pigment)
Zinc Oxide (pigment)
36,27
35
37
21


X
X
X
X
X
X
 27— nickel sulfate
 28— nitric acid mist
 29—phosphorus
 30— phosphoric acid mist
 31— phosphorus pentasulfide
 32— phosphorus trichloride
 33— sulfuric acid mist
 34—sulfur trioxide
 35— zinc chromate
 36— zinc chloride fume
 37— zinc oxide fume
 38—iodine
 39— hydrazine
 40— iron oxide
                Source Key
                A —converter
                B — absorption tower
                C — concentrator
                D —desulfurizer
                E — reformer
                F — neutralizer
                G—kiln
                H — reactor
                I  —crystallizer
                J — compressor and pump seals
                K — storage tank vents
                L —dryer
                M— leaching tanks
                N—filter
                0 — flakers
                P —milling/grinding/crushing
                Q — product handling and
                     packaging
                R — cooler (cooling tower,
                    condenser)
                S —pressure relief valves
                T — raw material unloading
                U — purification
                V — calciner
                W— hot well
                X — no information
 a References 6, 13, 20, 22, 23, 24, 25, 27, 28, 29, 30, 33, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51.
                                                                                         13

-------
Table 2-5. Potential HAP's and Emission Sources for the Chemical Products Industry

Source Category

Carbon Black

Charcoal
Explosives
Fertilizers
Paint and Varnish

Pharmaceutical

Plastics


Printing Ink

Pesticides


Soap and Detergents
Synthetic Fibers



Synthetic Rubber



Pollutant Key
1 — arsenic
2 — acrolein
3 — acrylonitrile
4 — acetic acid
5 — boron
6 — barium
7— beryllium
8 — benzene
9 — cresols
10 — cadmium
11 — chromium
12 — chloroprene
13 — caprolactum
14 — carbon disulfide
15 — carbonyl sulfide
16 — carbon tetrachloride
77— chloroform
18 — dichlorobenzene
19 — dimethylformamide
20 — dimethylamine
21— ethylene
22— ethylene dichloride
23— formaldehyde
24 — hydrogen sulfide
25 — hexachlorocyclopentadiene
26— hydrogen fluoride
27— ketones
28 — mercury
29 — manganese
30— methanol
Potential HAP's3
Organic Inorganic
Vapor Particulate Vapor Particulate
14,15,21 41 1 1,7,10,11
16,24 28,29,37
4,23,30 41
23
23,26,44 49
16,22, 6,28,43,48
31,46
17,18,31, 28
34,46
23,33,35,
39,42,50,
51,52
2, 16,27,
42,46,45
9,16,17,18, 28
20,25,32,
36,39,47
8,34 41 1 1,5
3,8,13,14
19,23,32,
38,40,42,
46,24
3,12,18,20, 41
22,33,34,
35,36,
46,49

31 — methyl chloroform
(1,1,1-trichloroethane)
32 — maleic anhydride
33— 1,3- butadiene
34 — morpholine
35 — metnylene chloride
36 — nitrosomines
37— nickel
38 — perchloroethylene
39 — phosgene
40— phthalic anhydride
41 — polycyclic organic matter
42 — phenol
43— selenium
44 — silicontetrafluoride
45 — terpenes
46— toluene
47 — xylene
48— zinc
49 — ammonia
50 — vinyl chloride
51 — toluene diisocyanate
52 — pyridine







Potential Emission Sources
Process Process Area
Point Fugitive Fugitive
B,H G,K,L 1

E
A,C,H K
D,H,R,S,V K,T
N,0 L

A,H,U,W G,L F

A,P,V K,L F,l


Q

A,H,O,X G F,l


M,N,0 K,L
A,H,J,O,U, G,K 1
v,x,z


A,H,0,P,X,Z Y F



Source Key
A — reactor
B — furnace
C — concentrator
D — neutralizer
E —kiln
F — compressor and pump seals; valves,
flanges, open ended lines, sampling lines
G — storage tank vents
H — dryer
1 — spills
J — spin cell or bath
K — product handling, finishing.
and packaging
L — raw material transport and unloading
M — spray dryer
N — kettle
0 — mixing tank (blend tank)
P — polymerization vessel
Q — cooking vessel
R — prill tower
S — granulator
T — screen
U —distillation
V — cooler (condenser)
W — crystallizer
X— filter
Y — milling/blending/compounding
Z —flash tank


References 6, 13, 21, 22, 23, 33, 52, 53, 54, 55, 56, 57, 58, 59, 60.




                          14

-------
Table 2-6.   Potential HAP's and Emission Sources for the Mineral Products Industry

                                            Potential HAP's"
                                             Potential Emission Sources
Source Category
Asbestos Products
Asphalt Batching Plants
Brick, Ceramic, and
Related Clay Products
Refractories
Cement Manufacture
Coal Cleaning (Dry Process)
Coal Cleaning (Wet Process)
Coal Conversion
Glass Fiber Manufacturing
Frit Manufacturing
Glass Manufacturing
Lime Manufacturing
Mercury Ore Processing
Mineral Wool Manufacturing
Perlite Manufacturing
Phosphate Rock Processing
Taconite Ore Processing
Organic Inorganic
Vapor Paniculate Vapor

2,8, 13 18
10,12
21,26
10,12
18 10,12
17,23


8, 19 18 4,23
27,28
13,19 19 6,20,22
12
1,4, 12,
14,26
15
15
13,19 12,23
12
6,20,21

Particulate
3,10,17

7,10,12
10,12
7,9, 10, 14
15,17,
24,25
22
1,5,6,7,9,
10,11,14,
16,17,20,
21,24
1,5,7,9,14
16,17,
20,29

12
1,5,6,12,
14,20,21
15
15

12
6,20,21
3
Process
Point

B
B,E,C
B,E
E

B,C
B,H
C,0
B,C
C
E,T
C
C,0
B,C
A,B,Q
C,Q
Process Area
Fugitive Fugitive
D,N I,L
F,J,M 1
D,F,IM I,L
D,F,N 1
F,G,N,S I,L
M,N,R I,L
M,N I,L
F,G,M,N I,L
D,F,G,N,P 1
S I,L
D,F,M,N 1
G,R,S I,L
G,N I,L
D,G,P I,L
G,M,N,S I,L
F,M,N,R I,L
F,M,N,R I,L
Pollutant Key
 1—arsenic
 2—aldehydes
 3— asbestos
 4— ammonia
 5— antimony
 6— barium
 7— beryllium
 8— benzene
 9—cadmium
10—chromium
11—copper
12— fluoride
13— formaldehyde
14— lead
15— mercury
16— manganese
77—nickel
18— polycyclic organic
    matter (POM)
19— phenol
20—selenium
21— boron
22—coal dust
23— hydrogen sulfide
24—zinc
25— iron
26— chlorine
27— cresols
28—toluene
29— phosphorus
Source Key
A —calciner
B —dryer
C —furnace
D — end-product forming and finishing
E —kiln
F — raw material preparation/mixing
G —cooling
H — reactor
I  — storage pile
J — saturator
L —mining operations
M — raw material handling/transport
N — raw material crusher/mill
0 — oven
P —resin application
Q —washers
R — screening
S —end-product handling/
     grinding/bagging
T — hydrator
a References 6, 13, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 45, 47, 64, 65, 66, 67, 68, 69, 70, 71, 73.
                                                                                       75

-------
Table 2-7.    Potential HAP's and Emission Sources for the Wood Products Industry

                                       Potential HAP's3
                                                        Potential Emission Sources
Source Category
Chemical Wood Pulping
Kraft pulp mill
Sulfite pulp mill
Neutral sulfite pulp mill
Plywood, Particleboard, Hardboard
Wood Preservative
Organic
Vapor Particulate

i e
h e
e
h,l,o,p
j,g,m,n
Inorganic
Vapor Particulate

k
f,k
k



a,b,c,d
a,b,c,d
a,c,d


Process
Point

A,B,C,D
A,B,C,
A,C,E
G

Process Area
Fugitive Fugitive




F
F
Pollutant Key
a —arsenic
b —asbestos
c —chromium
d — mercury
e —polycyclic organic matter (POM)
f —chlorine
g —chlorobenzene
h —formaldehyde
                i  —methyl mercaptan
                j  —dioxin
                k  — hydrogen sulfide
                I  —phenol
                m — pentachlorophenol
                n  —cresols
                o  — abietic acid
                p  — pinene
                                        Source Key
                                        A — recovery furnace
                                        B — digester
                                        C — blow tank
                                        D — lime kiln
                                        E —fluidized bed reactor
                                        F — resin and/or adhesive application
                                        G — dryer
"References 4, 13, 21, 22, 23, 33, 56, 73.
Table 2-8.    Potential HAP's for Petroleum Related Industries"
           (General Listing for Entire Source Category)

                   Potential HAP's
         Organic
                  Inorganic
     Vapor
Particulate
Vapor
Particulate
Parafins (CrC10)   Coke fines    Sulfides
Cycloparafins
Aromatics (e.g.,
  benzene, toluene
  xylene)
Phenols
Sulfur containing
  compounds (e.g.,
  mercaptans,
  thiophenes)
           (e.g., hydrogen
           sulfide, carbon
           disulfide,
           carbonyl sulfide)
           Ammonia
         Catalyst
          fines
"References 26, 75, 76, 77, 78, 79, 80.

composition of the two starting materials used in
these industries—crude oil and natural gas. Crude
oil is composed chiefly of hydrocarbons (paraffins,
naphthalenes,  and aromatics) with small amounts
of trace elements and organic compounds contain-
ing  sulfur,  nitrogen, and oxygen. Natural  gas is
largely saturated hydrocarbons (mainly methane).
The remainder may include nitrogen, carbon diox-
ide, hydrogen sulfide, and helium. Organic and in-
organic particulate emissions, such as coke fines or
catalyst  fines, may  be generated in  some pro-
cesses.

The emission sources within each of the petroleum
related industries are given in Table  2-10. Sources
of potential HAP emissions from  the oil and gas
production  industry include  blowouts during drill-
ing operations;  storage tank breathing  and filling
losses; wastewater treatment processes; and fugi-
tive leaks in valves, pumps, pipes, and vessels. In
the petroleum  refining industry, potential HAP
emission sources include  distillation/fractionating
columns, catalytic  cracking  units,  sulfur recovery
processes, storage tanks,  fugitives,  and combus-
tion  units (e.g.,  process heaters). Fugitive emis-
sions are a major source of emissions in this indus-
try. Emission sources in the basic petrochemicals
industry are similar to those from the  petroleum
refining  industry  and the  SOCMI  (see Section
2.2.3).

2.2.9 Combustion Sources
The fuel  combustion industry encompasses a large
number of combustion units generally used to pro-
duce  electricity, hot water, and process steam for
industrial plants; or to provide space  heating for
industrial, commercial, or  residential  buildings.
The combustion units may differ in size, configura-
tion,  and type of fuel  burned.  Coal, fuel oil, and
natural  gas are the major fossil fuels  burned, al-
though other fuels such as wood and  various waste
(e.g.,  waste oil) or byproduct fuels are burned in
relatively small quantities. Industrial applications
of both gasoline- and diesel-powered stationary in-
ternal combustion  units such as generators,
pumps,  and  well-drilling  equipment are also  in-
cluded in this category.

The waste incineration category includes combus-
tion processes whereby municipal solid wastes or
sewage treatment sludges are disposed. Table 2-11
presents the potential HAP's and the facility specif-
ic emission sources  for  the above combustion
sources.
                        16

-------
Table 2-9.
Process
Potential HAP's for Petroleum Refining Industries (26)
(Specific Listing for Petroleum Refining Segment)
                                                                       Potential HAP's
                                                     Organic
                                                                                  Inorganic
                                            Vapor
                                                 Paniculate
                                                                                      Vapor
Pollutant Key
a —maleic anhydride
b — benzoic acid
c —chlorides
d — ketones
e —aldehydes
f —heterocyclic compounds
     (e.g., pyridines)
g — benzene
h —toluene
i —xylene
j — phenols
k —organic compounds containing
     sulfur (sulfonates, sulfones)
I — cresol
m — inorganic sulfides
n — mercaptans
o — polynuclear compounds (benzo
     pyrene, anthracene, etc.)
                                           p — vanadium
                                           q — nickel
                                           r — lead
                                           s — sulfuric acid
                                           t — hydrogen sulfide
                                           u — ammonia
                                           v —carbon disulfide
                                           x — carbonyl sulfide
                                           y — cyanides
                                           z — chromates
                                           A — acetic acid
                                           B —formic acid
                                           C —methylethylamine
                                           D —diethylamine
                                           E —thiosulfide
                                           F — methyl mercaptan
Particulate
Crude Separation
Light Hydrocarbon Processing
Middle and Heavy
Distillate Processing
Residual Hydrocarbon Processing
Auxiliary Processes

a,b,d,e,f,g,h,i,j,k,l,m,o,
A,B,C,D,E,F,J
g,h,i,n,N,0,P
a,d,e,f,g,h,i,j, k,l,
F,J,K,0,P,S,T
a,d,e,f,g,h,i,j,k,l,n,
F,J,M,N,P,S,T
a,b,d,e,f,g,h,i,j,k,l,n,
A,B,C,D,J,K,M,T

0
R
o,R
o,R
o,R

c,m,t,u,
v,x,y,L
t,v
m,t,u,v,
x,y,L
m,s,t,u,
v,x,y,L
c,m,s,u,
y,L

P,I,Q,R
G,H,Q
P,q,G,H,
I,Q,U
P,q,G,H,
I,Q,U
P,q,r,z,
1

                                                                                                    G —cobalt
                                                                                                    H — molybdenum
                                                                                                    I  —zinc
                                                                                                    J —cresylic acid
                                                                                                    K —xylenols
                                                                                                    L —thiophenes
                                                                                                    M — thiophenol
                                                                                                    N — nickel carbonyl
                                                                                                    0 — tetraethyl lead
                                                                                                    P —cobalt carbonyl
                                                                                                    Q —catalyst fines
                                                                                                    R —coke fines
                                                                                                    S —formaldehyde
                                                                                                    T —aromatic amines
                                                                                                    U —copper
Table 2-10.    Emission Sources for Petroleum Related Industries
 Source Category
                                                                           Potential HAP Emission Sources
                                                     Process
                                                      Point
                                                                                      Process
                                                                                      Fugitive
   Area
 Fugitive
Oil and Gas Production
   Exploration, Site Preparation and Drilling
   Crude Processing
   Natural Gas Processing
   Secondary and Tertiary Recovery Techniques

Petroleum Refining Industry
   Crude Separation
   Light Hydrocarbon Processing
   Middle and Heavy Distillate Processing
   Residual Hydrocarbon Processing
   Auxiliary Processes
                                                        A
                                                        G
                                                      G,J,K
                                                        G
                                                      G,J,L
                                                       O,G
                                                     G,O,P,R
                                                    B,G,K,0,R
                                                        G
Source Key
A — blowout during drilling
B —visbreaker furnace
C —cuttings
D —drilling fluid
E — pipe leaks (due to corrosion)
F — wastewater disposal (process
     drain, blowdown, cooling water)
                                             G —flare, incinerator, process
                                                 heater, boiler
                                             H — storage, transfer, and handling
                                             I  —pumps, valves, compressors,
                                                 fittings, etc.
                                             J  —absorber
                                             K — process vent
                                                                                         C
                                                                                        F,H
                                                                                         H
                                                                                      F,H,M,N
                                                                                        F,H
                                                                                        F,H
                                                                                         H
                                                                                        F,H
   D,E
     I
    Q
     I
     I
Basic Petrochemicals Industry
Olefins Production
Butadiene Production
Benzene/Toluene/Xylene (BTX) Production
Naphthalene Production
Cresol/Cresylic Acids Production
Normal Paraffin Production

G,K,0
G,J,L,0,R
G,K,O,R
G,L,0
G,L
G,0

F,H
F,H,N
F,Q
F,H
F,H
F,H

1
1
1
1

1
                                                                         L —distillation/fractionation
                                                                         M — hotwells
                                                                         N — steam ejectors
                                                                         0 —catalyst regeneration
                                                                         P —evaporation
                                                                         Q — catalytic cracker
                                                                         R —stripper

                                                                             77

-------
Table 2-11.   Potential HAP's and Emission Sources for Combustion Sources
Source Category
Coal Combustion



Oil Combustion



Natural Gas Combustion
Gasoline Combustion
Diesel Combustion
Wood Combustion

Waste Oil Combustion

Municipal Refuse Incineration

Sewage Sludge Incineration

PCB Incineration

Potential
Organic
Vapor
14,25,28



14



14
12,14
12
3,4, 12,
74,25
7,12,21,
23,26
12

12

12,21
Particulate
19



19



19
12,19
12,19
12,19

12,19

12,19

12,19

12,19
HAP's
Inorganic
Vapor Particulate
1,2,8,9,13 1,2,5,6,8,9
17,27 10,11,15,
16, 18,20,
22,24
13,17,27 1,2,5,6,8,9,
10,11,15,
16,18,22,
24,29

17 15
6,18
27 16,20

6,8,9,
15,18
77,27 6,8,9,11,
15,16,18
17 1,6,8,9,
15,16,18
30
Potential
Process
Point
A,B



A,B,E



A,B,E,F,
G
G
A,B,C

A,B,D

D

D

D,B
Emission Sources
Process Area
Fugitive Fugitive
I H



















Pollutant Key
 7 — arsenic
 2— antimony
 3—acetaldehyde
 4— acetic acid
 5— barium
 6— beryllium
 7— benzene
 8—cadmium
 9—chromium
10—cobalt
77 — copper
72—dioxin
73—fluoride
14—formaldehyde
75—lead
16— manganese
77— mercury
18— nickel
19— polycyclic organic matter (POM)
20—phosphorus
27 — polychlorinated biphenyls (PCB)
22— radionuclides
23—trichloroethy lene
24—zinc
25— phenol
26— ethyl benzene
27— chlorine
28— pyridine
29—vanadium
30—dibenzofuran
Source Key
A —furnace
B —boiler
C —woodstove/fireplace
D — incinerator
E — gas turbine
F — reciprocating engine
G — industrial engine and/or equipment
H — coal storage pile
I  —ash handling system
a References 6, 13, 20, 21, 23, 24, 25, 26, 27, 28, 29, 30, 33, 34, 35, 53, 72, 81, 82, 83, 84 .
2.3  Identification of Key Emission Stream
     Properties
This section identifies the emission stream physical
properties needed to select the appropriate control
technique(s) and to size the control device(s) for
each identified HAP emission stream generated by
a process source, be it either a process point source
or a process fugitive source. Design and costing
techniques for area fugitive emission control meth-
odologies are outside the scope of this handbook;
however, control techniques for vapor emissions
and particulate emissions from  area fugitive
sources are discussed in Sections 3.2.4 and 3.3.2,
respectively.
The actual/estimated  values for the process emis-
sion stream properties should be obtained from the
owner/operator or from available literature if the
owner/operator cannot  supply the necessary data.
                   The values obtained are used in conjunction with
                   the guidelines given in Chapter 3 to perform the
                   control technique selection process. Table 2-12 lists
                   the required information for organic vapor emis-
                   sions, Table 2-13 for inorganic vapor  emissions,
                   and Table 2-14 for particulate emissions.  After ob-
                   taining the values for the key physical properties
                   for each HAP emission stream, record the data on
                   the HAP Emission Stream Data Form found in Ap-
                   pendix C.1 (see Figure 2-1).

                   There will be occasions when it would  be prudent
                   for the owner/operator to combine similar emis-
                   sion streams. For example, if two or more  emission
                   streams require  the use  of the same control tech-
                   nique,  it will normally be more cost effective to
                   combine the streams and use just one  control de-
                   vice as opposed to using a control device for each
                   separate emission stream. If the owner/operator
                       18

-------
decides to combine emission  streams, Appendix
B.1 provides calculation procedures to determine
the key effluent properties of combined  emission
streams.
                          Table 2-13.    Key Properties for Inorganic Vapor Emissions
                  Example Case
 The emission stream from the coating oven has
 been  identified  as containing  organic vapors.
 Therefore, the emission properties required are
 listed in Table 2-12. The necessary information
 was provided by the owner/operator; a source
 test was performed for a very similar operation
 at another plant owned by his company.  The
 information is recorded on the worksheet  pre-
 sented in Appendix  C.1. The source test  pro-
 vided the following data (see Figure 2-1):
  HAP content
  Organic content

  Moisture content
  Halogen content
  Metal content
  Temperature
  Pressure
  Flow rate
  Heat content
  Oxygen content
  HAP molecular weight
  HAP vapor pressure
  HAP solubility
73 ppm (vol) toluene
44 ppm (vol) methane
 4 ppm (vol) other
2% (vol)
None
None
120°F
Atmospheric
15,000 scfm (max)
0.4 Btu/scf
20.6% (vol)
92 Ib/lb-mole
28.4 mm Hg @ 77°F
Insoluble in water
                          Emission Stream Properties
                          (Preferred units of measure)
                                 HAP Properties3
                          HAP Contentb (ppm by volume)
                          Moisture Content (% by volume)
                          Halogen/Metal Content (yes or no)
                          Flow Rate (scfm)
                          Temperature (°F)
                          Pressure (mm Hg)
                                 Molecular Weight
                                 Vapor Pressure
                                 Solubility (graph)
                                 Adsorptive Properties
                                  (isotherm plot)
                          "These properties pertain  to the specific HAP or mixture of
                           HAP's in the emission stream.
                          bPrimary properties that affect control technique selection.
Table 2-14.   Key Properties for Paniculate Emissions

Emission Stream Properties              HAP Properties3
(Preferred units of measure)
HAP Content (% by mass)
Paniculate Contentb (Ib/acf)
Moisture Content (% by volume)
S03 Content (ppm by volume)
Flow Rate (acfm)
Temperature (°F)
Particle Mean Diameter0 (
Drift Velocity0 (ft/sec)
None
"These properties pertain to the specific  HAP or mixture of
 HAP's in the emission stream.
bData include total particulate loading and principle paniculate
 constituent.
These properties are necessary only for specific control tech-
 niques.
Table 2-12.    Key Properties for Organic Vapor Emissions

Emission Stream Properties          HAP Properties3
(Preferred units of measure)
HAP Content (ppm by volume)
Organic Content6 (ppm by volume)
Heat Content0 (Btu/scf)
Oxygen Content (% by volume)
Moisture Content (% by volume)
Halogen/Metal Content (yes or no)
Flow Rate (scfm)
Temperature (°F)
Pressure (mm Hg)
    Molecular Weight
    Vapor Pressure
    Solubility (graph)
    Adsorptive Properties
      (isotherm plot)
3These  properties pertain to the specific HAP or mixture of
 HAP's in the emission stream.
bPrimary properties that affect control technique selection. Or-
 ganic content is defined as organic emission stream combusti-
 bles less HAP emission stream combustibles.
°Heat content is determined from HAP/Organic Content (see
 Appendix B.1 for calculation procedures).
2.4  References
 1. U.S. EPA.  Evaluation of Control Technologies
    for Hazardous Air  Pollutants.  EPA-600/7-86-
    009 (NTIS PB 86-167/038/AS). October 1985.

 2. National Paint and Coatings Association. Sec-
    tion III: Paint and Coatings Markets. Table A-6,
    Estimated  Consumption of Solvents in  Paints
    and Coatings, by Market -  1981.  pp. 208-209.
    (no date).

 3. U.S. EPA.  Organic Solvent Cleaners - Back-
    ground Information for Proposed  Standard
    (Draft). EPA-450/2-78-045a. October 1979.

 4. U.S. EPA. End  Use of Solvents Containing
    Volatile Organic Compounds.  EPA-450/3-79-
    032. May 1979.

 5. U.S. EPA. Source Assessment: Solvent Evapo-
    ration  - Degreasing Operations. EPA-600/2-79-
    019f. August 1979.
                                                                             19

-------
 6. U.S. EPA. Compilation of Air Pollutant Emis-
   sion Sources. Third Edition: Supplements 1-15.
   AP-42. January 1984.

 7. U.S.  EPA.  Guidance  for Lowest Achievable
   Emission  Rates for  18 Major  Stationary
   Sources of Particulates, Nitrogen Oxides, Sul-
   fur Dioxide, or Volatile Organic Compounds.
   EPA-450/3-79-024. April 1979.

 8. U.S.  EPA.  Control of Volatile Organic Emis-
   sions from Existing Stationary Sources - Vol-
   ume VI: Surface Coating of Miscellaneous Met-
   al Parts and Products. EPA-450/2-78-015. June
   1978.

 9. U.S.  EPA.  Control of Volatile Organic Emis-
   sions from Existing Stationary Sources - Vol-
   ume II: Surface Coating of Cans, Coils, Paper,
   Fabrics, Automobiles, and Light Duty  Trucks.
   EPA-450/2-77-008. May 1977.

10. U.S.  EPA.  Control of  Volatile Organic Com-
   pound Emissions from Large Petroleum Dry
   Cleaners. EPA-450/3-82-009. September 1982.

11. U.S. EPA. Pressure Sensitive Tape and Label
   Surface Coating Industry -  Background Infor-
   mation for Proposed Standards. EPA-450/3-80-
   003a. September 1980.

12. U.S. EPA. Hazardous/Toxic  Air Pollutant Con-
   trol Technology: A  Literature Review. EPA-
   600/2-84-194. December 1984.

13. U.S. EPA. Nonindustrial Sources of Toxic Sub-
   stance Emissions  and Their Applicability to
   Source Receptor Modeling. Draft Report, EPA
   Contract No. 68-02-3509, Task No. 42. July 27,
   1983.

14. U.S. EPA. Control Technique Guidelines for the
   Control of  Volatile Organic  Emissions from
   Wood Furniture Coating (Draft). April 1979.

15. U.S. EPA. Flexible Vinyl Coating and Printing
   Operations - Background Information for Pro-
   posed Standards. EPA-450/3-81-016a. January
   1983.

16. U.S.  EPA. Background Information for New
   Source Performance Standards: Primary Cop-
   per, Zinc, and Lead Smelters - Volume 1: Pro-
   posed Standards. EPA-450/2-74-002a. October
   1974.

17. U.S.  EPA.  Background Information for Stan-
   dards of Performance: Electric Submerged Arc
   Furnaces  for Production  of Ferroalloys - Vol-
   ume 1:  Proposed  Standards.  EPA-450/2-74-
   018a. October 1974.

18. U.S.  EPA. Control Techniques for Volatile Or-
   ganic Compound Emissions  from Stationary
   Sources- Third Edition (Draft). April 1985.
19.  U.S. EPA. A Method for Characterization and
    Quantification of Fugitive Lead Emissions from
    Secondary Lead Smelters, Ferroalloy Plants,
    and Gray  Iron Foundries. EPA-450/3-78-003.
    January 1978.

20.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Chromium. EPA-450/4-
    84-007g. July 1984.

21.  U.S. EPA. A Survey of Emissions and Controls
    for Hazardous and Other  Pollutants. EPA-R4-
    73-021. February 1973.

22.  U.S. EPA. Industrial Sources of Hazardous Air
    Pollutants (Draft). September 1983.

23.  U.S. EPA. Source Assessment: Noncriteria Pol-
    lutant Emissions (1978 Update). EPA-600/2-78-
    004T. July  1978.

24.  U.S. EPA. Locating and Estimating Air Emis-
    sions from  Sources of Manganese (Draft). Sep-
    tember 1984.

25.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Nickel. EPA-450/4-84-
    007F. March 1984.

26.  U.S.  EPA. Potentially Hazardous Emissions
    from the Extraction and Processing of Coal and
    Oil. EPA-650/2-75-038. April 1975.

27.  U.S. EPA. Review of National Emission Stan-
    dards for Mercury. EPA-450/3-84-01. December
    1984.

28.  U.S. EPA. Status Assessment of Toxic Chemi-
    cals: Lead.  EPA-600/2-79-210h. December
    1979.

29.  U.S. EPA. Status Assessment of Toxic Chemi-
    cals:  Mercury. EPA-600/2-79-210J.  December
    1979.

30.  U.S. EPA.  Sources of  Copper Air Emissions.
    EPA-600/2-85-046. April 1985.

31.  U.S. EPA. The Use and Fate of Lubricants, Oils,
    Greases and Hydraulic Fluids in The Iron and
    Steel Industry. EPA-600/2-78-101. May 1978.

32.  U.S. EPA. Rubber Tire Manufacturing Industry -
    Background Information  for Proposed Stan-
    dards. EPA-450/3-81-008a.  July 1981.

33.  U.S. EPA.  Human  Exposure to Atmospheric
    Concentrations of Selected Chemicals. EPA
    Contract No.  68-02-3066. February 1982.

34.  U.S. EPA. Survey of Cadmium  Emission
    Sources. EPA-450/3-81-013. September 1981.

35.  U.S. EPA. Source Category Survey: Secondary
    Zinc Smelting and Refinery Industry. EPA-
    450/3-80-012. May 1980.
                     20

-------
36.  U.S. EPA.XWr Oxidation Processes in Synthetic
    Organic  Chemical Manufacturing  Industry -
    Background Information  for Proposed Stan-
    dards. EPA-450/3-82-001a. October 1983.

37.  U.S. EPA. Reactor Processes in Synthetic Or-
    ganic Chemical Manufacturing  - Background
    Information for Proposed Standards  (Draft).
    October  1984.
38.  U.S. EPA. VOC Emissions from Volatile Organic
    Liquid Storage  Tanks — Background Informa-
    tion for Proposed Standards (Draft). EPA-450/3-
    81-003. July 1984.

39.  U.S. EPA. VOC Fugitive Emissions in Synthetic
    Organic  Chemicals Manufacturing Industry -
    Background Information for Promulgated Stan-
    dards. EPA-450/3-80-033b. June  1982.

40.  U.S. EPA. Distillation Operations in Synthetic
    Organic Chemical Manufacturing - Background
    Information  for Proposed Standards.  EPA-
    450/3-83-005a. December 1983.

41.  U.S.  EPA. Organic Chemical  Manufacturing
    Volume  6: Selected Processes. EPA-450/3-80-
    028a. December 1980.

42.  U.S. EPA. Source Category: Ammonia Manu-
    facturing Industry. EPA-450/3-80-014.  August
    1980.

43.  U.S. EPA. Source Assessment: Ammonium Ni-
    trate  Production. EPA-600/2-77-1071. Septem-
    ber 1977.

44.  U.S.  EPA. Ammonium Sulfate  Manufacture -
    Background  Information for Proposed Stan-
    dards. EPA-450/3-79-034a. December 1979.

45.  U.S. EPA. Preliminary Study of Sources of Inor-
    ganic Arsenic. EPA-450/5-82-005. August 1982.

46.  U.S. EPA. Source Assessment:  Major  Barium
    Chemicals. EPA-600/2-78-004b. March 1978.
47.  U.S.  EPA. Emission Factors  for  Trace Sub-
    stances. EPA-450/2-73-001. December 1973.

48.  U.S. EPA. Review of New Source Performance
    Standards for Nitric Acid Plants. EPA-450/3-84-
    011. April 1984.

49.  U.S. EPA. Sodium Carbonate Industry - Back-
    ground  Information for Proposed  Standards.
    EPA-450/3-80-029a. August 1980.

50.  U.S. EPA. Industrial Process Profiles for Envi-
    ronmental Use: Sulfur, Sulfur Oxides and Sul-
    furicAcid. EPA-600/2-77-023w. February 1977.

51.  U.S. EPA. Final Guideline Document:  Control
    of Sulfuric Acid Mist Emissions from Sulfuric
    Acid Production Plants. EPA-450/2-77-019. Sep-
    tember 1977.
52.  U.S. EPA. Source Assessment: Charcoal Manu-
    facturing. EPA-600/2-78-004z. December 1978.

53.  U.S. EPA. Locating and Estimating Air Emis-
    sions from  Sources of Formaldehyde.  EPA-
    450/4-84-007e. March 1984.

54.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Chloroform. EPA-450/4-
    84-007c. March 1984.

55.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Carbon  Tetrachloride.
    EPA-450/4-84-007b. March 1984.

56.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Chlorobenzenes (Draft).
    September 1984.

57.  U.S. EPA. Plastics and Resins Industry - Indus-
    trial Process Profiles for Environmental Use.
    EPA-600/2-77-023J. February 1977.

58.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources of Phosgene (Draft). Sep-
    tember 1984.

59.  U.S. EPA. Locating and Estimating Air Emis-
    sions from Sources ofAcrylonitrile. EPA-450/4-
    84-007a. March 1984.

60.  U.S. EPA. Locating and Estimating Air Emis-
    sions from  Sources of Ethylene Dichloride.
    EPA-450/4-84-007d. March 1984.

61.  U.S. EPA. Asphalt Roofing Manufacturing In-
    dustry - Background Information for Proposed
    Standards (Draft). EPA-450/3-80-021a. June
    1980.

62.  U.S. EPA. Trace Pollutant Emissions from the
    Processing of Nonmetallic Ores. EPA-650/2-74-
    122. November 1974.

63.  U.S. EPA. Source Category Survey: Refractory
    Industry. EPA-450/3-80-006. March 1980.
64.  U.S. EPA. A Review of Standards  of Perfor-
    mance for New Stationary Sources - Portland
    Cement Industry.  EPA-450/3-79-012.  March
    1979.

65.  U.S. EPA. Background Information for Stan-
    dards of Performance: Coal Preparation Plants
    Volume I: Proposed Standards.  EPA-450/2-74-
    021 a. October 1974.
66.  U.S. EPA. Glass Manufacturing Plants, Back-
    ground Information: Proposed  Standards  of
    Performance (Draft). EPA-450/3-79-005a. June
    1979.

67.  U.S. EPA. Wool Fiberglass Insulation Manufac-
    turing  Industry - Background Information for
    Proposed Standards (Draft).  EPA-450/3-83-
    002A. December 1983.
                                                                    21

-------
68.  U.S. EPA. Standards Support and Environmen-
    tal Impact Statement Volume I: Proposed Stan-
    dards of Performance for Lime Manufacturing
    Plants. EPA-450/2-77-007a. April 1977.

69.  U.S. EPA. Final Standards Support and Envi-
    ronmental Impact Statement Volume II: Pro-
    mulgated Standards of Performance for Lime
    Manufacturing Plants. EPA-450/2-77-007b. Oc-
    tober 1977.

70.  U.S. EPA. Source  Category Survey: Mineral
    Wool Manufacturing Industry.  EPA-450/3-80-
    016. March 1980.

71.  U.S. EPA. Source Category Survey: Perlite In-
    dustry. EPA-450/3-80-005. May 1980.

72.  U.S. EPA. Radionuclides - Background Informa-
    tion Document for Final Rules.  Volume I. EPA-
    520/1-84-022-1. October 1984.

73.  Standards Support and Environmental Impact
    Statement for the Iron Ore Benefication Indus-
    try (Draft). Battelle Columbus Laboratories. De-
    cember 1976.

74.  U.S. EPA. Kraft Pulping - Control of TRS Emis-
    sions from Existing Mills. EPA-450/2-78-003b.
    March 1979.

75.  U.S. EPA. Industrial Process Profiles for Envi-
    ronmental Use: Chapter 2. Oil and Gas Produc-
    tion Industry.  EPA-600/2-77-023b.  February
    1977.

76.  U.S. EPA. Industrial Process Profiles for Envi-
    ronmental Use: Chapter 3. Petroleum Refining
    Industry. EPA-600/2-77-023c. January 1977.
77.  U.S. EPA. Industrial Process Profiles for Envi-
    ronmental Use: Chapter 5. Basic Petrochemi-
    cals Industry. EPA-600/2-77-023e. January
    1977.

78.  U.S. EPA. VOC Fugitive Emissions in Petroleum
    Refining Industry - Background Information for
    Proposed Standards. EPA-450/3-81-015a.  No-
    vember 1982.

79.  U.S. EPA. VOC Species Data Manual, Second
    Edition. EPA-450/4-80-115. July 1980.

80.  U.S.  EPA. Bulk Gasoline Terminals - Back-
    ground Information  for Proposed Standards
    (Draft). EPA-450/3-80-038a. December 1980.

81.  U.S.  EPA. Air Toxics Emission Patterns  and
    Trends - Final Report. EPA Contract No. 68-02-
    3513, Task 46. July 1984.

82.  U.S.  EPA. Hazardous Emission Characteriza-
    tion of Utility Boilers. EPA-600/2-75-066.  July
    1975.

83.  U.S.  EPA. Thermal Conversion of Municipal
    Wastewater Sludge Phase II: Study of Heavy
    Metal Emissions.  EPA-600/2-81-203. Septem-
    ber 1981.

84.  U.S.  EPA. Locating and Estimating Air Emis-
    sions from Sources of Polychlorinated Biphen-
    yls (Draft). November 1984.
                     22

-------
                                           Chapter 3
                                   Control Device Selection
3.1  Background
This chapter presents guidelines that will enable
the user to select the control technique(s) that can
be used to control HAP's. The control techniques
that can be applied to control HAP emissions from
a specific emission source will depend on the emis-
sion source characteristics and HAP characteristics.
Therefore, Section 3.2, Vapor Emissions Control,
and  Section 3.3,  Particulate Emissions Control,
each pertain to specific HAP groups. The discus-
sion of control technique selection within each sec-
tion is according to type  of HAP (organic or inor-
ganic) and emission source (point, process fugitive,
or area fugitive).
In the following sections, guidelines for selecting
controls for point sources are discussed in detail.
Point sources are typically controlled  by add-on
control devices. For each control technique, ranges
of applicability with respect  to  emission stream
characteristics, HAP characteristics,  performance
levels (e.g., removal  efficiency), and other consid-
erations that are important in  control device selec-
tion are described in  detail.
Work practices, including equipment modifica-
tions, play a key role in reducing emissions from
process fugitive and area fugitive sources. These
sources can also be controlled by add-on control
devices if the emissions can be captured by hood-
ing or enclosure or collected by closed vent sys-
tems and then transferred to a control device. Note
that the overall performance of the control system
will then be dependent on both the capture efficien-
cy of the  fugitive emissions and the efficiency of
the control device.
To  illustrate the control device selection process,
several  emission stream scenarios are presented
throughout this chapter. The emission stream from
the paper coating drying oven introduced in Chap-
ter  2 is  one of the scenarios presented. The data
necessary for control device selection are recorded
on the HAP Emission Stream Data Form (see Figure
2-1).
3.2  Vapor Emissions Control
3.2.1  Control Techniques for Organic Vapor
      Emissions from Point Sources
The most frequent approach to point source con-
trol is the application  of add-on  control devices.
These devices  can  be  of two types:  combustion
and recovery.  Applicable combustion devices are
thermal incinerators, catalytic incinerators, flares,
and boilers/process  heaters. Applicable  recovery
devices include condensers, adsorbers,  and ab-
sorbers.  The combustion  devices are the more
commonly applied control  devices, since they are
capable of high removal (i.e., destruction) efficien-
cies for almost any type of HAP  (organic vapor).
The removal efficiencies of the recovery techniques
generally depend  on the physical and chemical
characteristics of the HAP under consideration.
Applicability  of  the control techniques  depends
more on the individual emission stream under con-
sideration than on  the particular source  category
(e.g., degreasing vs. surface coating in solvent us-
age operations source category). Thus, selection of
applicable control  techniques  for point source
emissions is made  on the basis of stream-specific
characteristics and desired control efficiency. Table
3-1 identifies the key emission stream characteris-
tics and HAP characteristics that  affect the applica-
bility of each control technique and presents limit-
ing  values for each  of these characteristics.
Matching the specific characteristics of the stream
under consideration with the corresponding values
in Table 3-1  will help  the user  to identify those
techniques that can potentially be used to control
the emission stream. The list of potentially applica-
ble control techniques  will then  be narrowed fur-
ther depending on the  capability of the applicable
control  devices  to achieve the required  perfor-
mance levels. Figure 3-1  identifies the expected
emission reduction from the application of each
control technique  on the basis  of the total VOC
concentration  in the emission stream. Very little
data  regarding control  device removal  efficiency
                                               23

-------
Table 3-1.     Key Emission Stream and HAP Characteristics for Selecting Control Techniques for Organic Vapors from Point Sources

                            Emission Stream Characteristics                            HAP Characteristics3
Control Device
Thermal
Incinerator
Catalytic
Incinerator
Flare
Boiler/
Process Heater9
Carbon
Adsorber



Absorber


Condenser


HAP/Organics
Contents6
(ppmv)
>20;
(<25% of LELC)
50-10,000;
(<25% of LELC)



1,000-10,000
(<25% of LELC)



250-10,000


>5,000


Heat Moisture
Content Content Flow Rate
(Btu/scf) (%) (scfm)

<100,000d

< 100,000
>300e <2,000,000f

>150h Steady

50%' 300-100,000



1,000-100,000


<2,000


Molecular
Weight Vapor
Temp. (Ib/lb- Pressure Adsorptive
(°F) mole) Solubility (mm Hg) Properties







Must be able
100-200 45-130 to adsorb on/
desorb from
available
adsorbents
Must be readily
soluble in water
or other solvents
>10(at
room temp-
erature)
"Refers to the characteristics of the individual HAP if a single HAP is present and to that of the HAP mixture if a mixture of HAP's is
  present.
bDetermined from HAP/hydrocarbon content.
°For emission streams that are mixtures of air and VOC; in some cases, the LELcan be increased to 40 to 50% with proper monitoring
  and control (see Section 4.2 for definition of LEL).
dFor packaged units; multiple-package or custom-made units can handle larger flows.
6 Based on EPA's guidelines for 98% destruction efficiency.
'Units: Ib/hr. Source: Reference 14.
9 Applicable if such a unit is already available on site.
hTotal heat content.
'Relative humidity.
Figure 3-1.   Percent reduction ranges for add-on control devices.
Thermal Incineration
T »95% T
T T
Catalytic Incineration
T »90% T
T * T
Carbon Adsorption
Absorption
Condensation
i i i
99%

95%

T 50%
T
T »90%T
T " T
I,,,

i i i

T
T 95% _ 99% _
T T * T
95% ^_ 98% T
T " T
50% 80% _ 95%
T * T *

10
          20
50        100       200   300      500       1,000     2,000  3,000    5,000     10,000    20,000

                     Inlet Concentration, HAPe (ppmv)

   24

-------
for specific HAP's are available. Therefore, without
actual  source test  data  for a  specific  emission
stream and control system, HAP removal efficiency
is  assumed to  equal  total  volatile  organic com-
pound (VOC) removal efficiency.


3.2.1.1 Thermal Incinerators
Thermal  incinerators are used to control  a wide
variety of continuous emission streams containing
VOC's. Compared to the other techniques, thermal
incineration is broadly applicable; that is, it is much
less  dependent on  HAP characteristics and emis-
sion  stream characteristics. Destruction efficiencies
up to 99+  percent are achievable with thermal in-
cineration. Although  they accommodate minor
fluctuations in  flow, thermal  incinerators are not
well  suited to  streams with highly variable flow
because the reduced residence time and poor mix-
ing during  increased flow conditions decrease the
completeness of combustion. This causes the com-
bustion chamber temperature to fall, thus decreas-
ing the destruction efficiency.

Thermal  incineration is typically applied to emis-
sion  streams that are  dilute mixtures of VOC  and
air.  In such  cases,  due to safety considerations,
concentration of the VOC's is generally limited by
insurance companies to 25 percent of the LEL (low-
er explosive limit)  for the VOC in  question (see
Section 4.2.2 for more details). Thus, if the VOC
concentration is high, dilution may be required.

When emission streams treated by thermal inciner-
ation are dilute (i.e., low heat content), supplemen-
tary  fuel  is required to maintain the desired com-
bustion  temperatures. Supplementary  fuel
requirements may  be reduced by  recovering the
energy contained in the hot flue gases from the
incinerator. For emission streams with high heat
contents (e.g., > 150 Btu/scf), the possibility of us-
ing the emission stream  as fuel gas should  be con-
sidered.
Packaged single unit thermal incinerators are avail-
able in many sizes to control emission streams with
flow rates from a few hundred up to about 100,000
scfm.

3.2.1.2 Catalytic Incinerators
Catalytic incinerators are similar to thermal inciner-
ators in design and operation except that the for-
mer  employ a catalyst to enhance the reaction rate.
Since the catalyst allows the reaction to take place
at lower temperatures, significant fuel savings are
possible with catalytic incineration.

Catalytic  incineration is not as broadly applied as
thermal incineration since performance of catalytic
incinerators is more sensitive to pollutant charac-
teristics and process conditions than is thermal in-
cinerator performance. Materials such as phospho-
rus,  bismuth, lead,  arsenic, antimony, mercury,
iron  oxide, tin, zinc, sulfur, and  halogens in the
emission stream  can poison the  catalyst  and se-
verely affect  its performance. (Note:  Some cata-
lysts can handle emission streams containing halo-
genated compounds.) Liquid or solid particles that
deposit on the catalyst and  form a coating also
reduce the catalyst's activity by preventing contact
between the VOC's and the catalyst surface. Cata-
lyst life is limited by thermal aging and by loss of
active sites by erosion, attrition, and vaporization.
With proper operating temperatures and adequate
temperature control, these processes are normally
slow, and satisfactory performance can be main-
tained for 3 to  5  years before replacement of the
catalyst is necessary.

Catalytic incineration is generally less expensive
than  thermal incineration  in  treating  emission
streams with low VOC concentrations. Emission
streams with  high VOC concentrations should not
be treated by catalytic incineration without dilution
since such streams may cause the catalyst bed to
overheat and lose its activity. Also, fluctuations in
the VOC content of the emission stream should be
kept to a minimum to prevent damage to the cata-
lyst.

Destruction efficiencies of up  to 95 percent of
HAP's are generally achieved with  catalytic inciner-
ation. Higher destruction efficiencies (99 percent)
are also achievable, but require larger catalyst vol-
umes and/or higher temperatures.

Catalytic incinerators have been applied to continu-
ous emission streams with flow rates  up to about
100,000 scfm.

3.2.1.3 Flares
Flares are commonly used for disposal of waste
gases during process upsets (e.g., start-up, shut-
down) and emergencies. They are basically safety
devices that are also used to destroy waste emis-
sion streams.

Flares can be used for controlling  almost any VOC
emission stream. They can be designed and oper-
ated to handle  fluctuations in emission VOC con-
tent, inerts content, and flow rate. There are several
different types  of flares including steam-assisted,
air-assisted, and  pressure head flares. Steam-as-
sisted flares are  very common and typically em-
ployed in cases where large volumes of waste gas-
es are released.  Air-assisted flares are  generally
used for moderate relief gas flows. Pressure head
flares are small; they are used in arrays of up to 100
individual flares. Normally, only a  few of the flares
operate.  The number  of flares operating is in-
creased as the gas flow increases.

Flaring  is generally considered a control option
when the heating value of  the emission stream
                                                                       25

-------
cannot be recovered because of uncertain or inter-
mittent flow as in process upsets or emergencies. If
the waste gas to be flared does not have sufficient
heating value to sustain combustion, auxiliary fuel
may be required.

Based  on studies conducted by  EPA, 98  percent
destruction efficiency  can  be achieved  by steam-
assisted flares when controlling emission streams
with heat contents greater than  300 Btu/scf.  De-
pending  on the type  of flare configuration (e.g.,
elevated  or ground flares), the capacity of flares to
treat waste gases can vary—up to about  100,000
Ib/hr for ground flares and 2 million Ib/hr or more
for elevated flares. The capacity of an array of pres-
sure head flares depends on the number of flares in
the array.

3.2.1.4 Boilers/Process Heaters
Existing boilers or  process heaters can be used to
control emission streams containing organic com-
pounds.  These are currently used as control  de-
vices for emission streams from several industries
(e.g., refinery operations, SOCMI reactor processes
and distillation operations, etc.)

Typically, emission streams are controlled in boil-
ers or  process heaters and used as supplemental
fuel  only if they  have  sufficient  heating  value
(greater than about 150 Btu/scf). In some instances,
emission  streams with high heat content may be
the main fuel to the process heater or boiler (e.g.,
process off-gas from  ethylbenzene/styrene manu-
facturing). Note that  emission streams with  low
heat content can also  be burned  in boilers or pro-
cess heaters when the flow rate of the emission
stream is small compared to the flow rate  of the
fuel/air mixture.

When used as emission  control devices, boilers or
process  heaters can provide destruction  efficien-
cies of greater than 98 percent at small capital cost
and  little or no fuel cost. In addition, near complete
recovery  of the emission stream heat content is
possible.

There  are some limitations in the  application of
boilers or process  heaters as emission control  de-
vices. Since these combustion devices are essential
to the operation of a plant, only those emission
streams  that will not  reduce their performance or
reliability can be controlled using these  devices.
Variations in  emission  stream  flow rate  and/or
heating  value  could  adversely affect the perfor-
mance of a boiler or  process heater. By lowering
furnace temperatures, emission streams with large
flow rates and low heating values can cause incom-
plete combustion and reduce heat output. The  per-
formance and  reliability of the process heater or
boiler may also be affected by the presence  of cor-
rosive compounds in the emission  stream; such
streams are usually not destroyed in these devices.
                     26
3.2.1.5 Carbon Adsorbers
Carbon  adsorption  is commonly  employed  as  a
pollution control and/or a solvent recovery tech-
nique. It is applied to dilute mixtures of VOC and
air. Removal efficiencies of 95 to 99 percent can be
achieved using carbon adsorption. The maximum
practical inlet concentration is usually about 10,000
ppmv. The inlet concentrations are typically limited
by the adsorption  capacity of the carbon  bed or
safety problems  posed by high  bed temperatures
produced  by heat  of adsorption and presence of
flammable vapors. Outlet concentrations of 50 to
100 ppmv can be routinely achieved with state-of-
the-art systems;  concentrations  as low as 10 to 20
ppmv can be achieved with some compounds. In
contrast to incineration methods whereby the
VOC's are destroyed, carbon adsorption provides a
favorable control alternative when the VOC's in the
emission stream  are valuable.

High molecular-weight compounds that are charac-
terized by low volatility are strongly adsorbed on
carbon.  The affinity of carbon for these compounds
makes it difficult to remove them during regenera-
tion of the carbon bed. Hence, carbon adsorption is
not applied to such compounds (i.e., boiling point
above 400°F; molecular weight greater than about
130).  Highly volatile  materials  (i.e.,  molecular
weight less than about 45) do not adsorb readily on
carbon; therefore,  adsorption is not typically used
for controlling emission streams  containing  such
compounds.

Carbon  adsorption is relatively  sensitive to emis-
sion stream conditions. The presence of liquid or
solid particles, high boiling organics, or polymeri-
zable substances may require pretreatment proce-
dures such as filtration. Dehumidification is neces-
sary if the emission stream has a high humidity
(relative humidity > 50 percent) and cooling may
be required if the emission stream temperature ex-
ceeds 120° - 1 SOT.

To prevent excessive bed temperatures resulting
from the exothermic adsorption process and oxida-
tion  reactions in the bed, concentrations higher
than 10,000 ppmv must frequently be reduced. This
is usually done by condensation or dilution of the
emission stream ahead of the adsorption step. Ex-
othermic reactions may also occur if incompatible
solvents are mixed in the bed, leading to polymer-
ization.  If flammable vapors are present, the VOC
concentrations may be limited by insurance  com-
panies to  less than 25 percent of the LEL. If proper
controls and monitors are used,  LEL levels up to 40
to 50 percent may be allowed.

Packaged carbon adsorption systems are available
that can handle  emission streams with flow rates
from a few hundred to above 100,000 scfm.

-------
3.2.1.6 Absorbers (Scrubbers)
Absorption is widely used as a raw material and/or
a product recovery technique in separation and pu-
rification of gaseous streams containing high con-
centrations of VOC's. As an emission control tech-
nique, it is much more commonly  employed for
inorganic vapors (e.g., hydrogen sulfide, chlorides,
etc.) than for organic vapors. Using  absorption as
the primary control technique for organic vapor
HAP's is subject to several limitations and prob-
lems as discussed below.

The suitability of absorption for controlling organic
vapor emissions is determined by several factors;
most of these factors will  depend on the specific
HAP in question. For example, the most important
factor is the availability of a suitable solvent. The
pollutant in question should be readily soluble in
the solvent for effective absorption  rates and the
spent solvent should be easily regenerated or dis-
posed of in an environmentally acceptable manner.

Another factor that affects the suitability of absorp-
tion  for organic vapor emissions  control is  the
availability of vapor/liquid equilibrium data for the
specific HAP/solvent system in question. Such data
are necessary for design of absorber systems. For
uncommon HAP's, these data are not readily avail-
able.

Another consideration  involved in the application
of absorption as a control technique is disposal of
the absorber effluent (i.e., used solvent).  If the ab-
sorber effluent containing the organic compounds
is discharged to the  sewer,  pond, etc., the air pollu-
tion  problem is merely being transformed into a
water  pollution problem.  Hence, this  question
should be addressed (e.g.,  are there  chemical/phy-
sical/biological means for treating the specific efflu-
ent under consideration?).  In  solvent  recovery,
used  organic solvents are typically stripped  (re-
verse of absorption) and recycled to the absorber
for economic  reasons. However, in HAP control
applications, stripping  requirements will often  be
very expensive because the residual organic con-
centrations in the solvent  must be extremely low
for it to be suitable  for reuse. Also, if the VOC's in
the effluent from the absorber have appreciable
vapor pressure (e.g., > 0.1  mm Hg), the possibility
of VOC emissions to the  atmosphere should  be
considered.

In organic vapor HAP control applications, low out-
let concentrations will typically be required. Trying
to meet such requirements with absorption alone
will lead to impractically  tall absorption towers,
long contact times,  and high liquid-gas ratios that
may not be cost effective. Therefore,  absorbers will
generally be effective when they are used  in combi-
nation with other control devices such as inciner-
ators.
Removal efficiencies in excess of 99 percent can be
achieved with absorption.
3.2.1.7 Condensers
Condensers are widely used as raw material and/or
product recovery devices. They  are frequently ap-
plied  as preliminary air pollution control devices
for removing  VOC contaminants from  emission
streams prior to other control devices such as in-
cinerators, adsorbers, or absorbers.

Condensers are also used by themselves for con-
trolling  emission  streams  containing  high VOC
concentrations (usually > 5,000 ppmv).  In these
cases, removal efficiencies obtained by condensers
range from 50 to 90 percent. The removal efficiency
of a  condenser is dependent  on the  emission
stream characteristics including the nature  of the
HAP in question (vapor pressure-temperature rela-
tionship) and HAP concentration, and the type of
coolant used. Note that a condenser cannot lower
the inlet VOC concentration to  levels below the
saturation concentration (or vapor pressure) at the
coolant temperature. When water, the most com-
monly used coolant, is employed, the saturation
conditions  represent high  outlet concentrations.
For example, condenser outlet VOC concentrations
are often limited to above 10,000 to 20,000 ppmv
due to the saturation  conditions of  most  of the
organic compounds at the temperature of the cool-
ing water. Therefore, it is not possible for conden-
sation with water as the coolant  to achieve the low
outlet concentrations that would be  required in
HAP control applications.

Removal efficiencies above 90 percent can be
achieved if lower temperatures than those possible
with cooling water are  employed. These low tem-
peratures can  be obtained with coolants such as
chilled water,  brine solutions, or chlorofluorocar-
bons. However, for extremely low outlet HAP con-
centrations, condensation will usually be economi-
cally infeasible.

Depending on the type of condenser used, there
may be potential problems associated with the dis-
posal of the spent coolant. Therefore, using contact
condensers that generate such  effluents  for con-
trolling HAP emissions  is not recommended.


Flow rates up to about 2,000 scfm can be consid-
ered as representative of the typical range for con-
densers used  as emission  control devices. Con-
densers for emission streams with flow rates above
2,000 scfm  and containing high concentrations of
noncondensibles will require prohibitively large
heat transfer areas.
                                                                      27

-------
3.2.2 Control Techniques for Inorganic Vapor
     Emissions From Point Sources
Inorganic vapors make up only a small portion of
the total HAP's emitted to the atmosphere. Poten-
tial sources of the various  inorganic vapors found
in the atmosphere are discussed in Chapter 2. Inor-
ganic HAP vapors typically include gases such as
ammonia, hydrogen sulfide, carbonyl sulfide, car-
bon disulfide, metals with hydride and carbonyl
complexes, chlorides, oxychlorides, and cyanides.

In many cases, although the inorganic HAP's are
emitted as vapors at the emission source, they may
condense when passing through various ducts and
form particulates. Prior to discharge to the atmo-
sphere, these  particulates  are typically controlled
by methods that will be discussed in Section 3.3. In
this section, the discussion will be based on control
techniques for HAP's that are emitted as vapors to
the atmosphere.

Only a  limited number of control  methods are ap-
plicable to  inorganic vapor emissions from point
sources. The two  most commonly used control
methods are absorption (scrubbing)  and adsorp-
tion. Absorption is the most widely used and ac-
cepted method for inorganic  vapor control. Al-
though combustion  can be used for some
inorganic HAP's  (e.g., hydrogen sulfide, carbonyl
sulfide, nickel carbonyl), typical combustion meth-
ods such as thermal and catalytic incineration are
generally not applied. In some cases, for example,
in controlling hydrogen sulfide emissions from gas
wells and gas processing, flares are used.

Applicability of absorption and adsorption as con-
trol methods depends on the individual emission
stream characteristics. The  removal  efficiencies
that can be achieved will be determined by the
physical and chemical properties of the HAP under
consideration. Other factors (e.g., waste disposal,
auxiliary equipment requirements), while not nec-
essarily affecting the technical feasibility of the
control device, may affect the decision to  use that
particular control method.

In the following two subsections, the applicability
of absorption  and adsorption for controlling  inor-
ganic vapor emissions will be discussed.


3.2.2.1  Absorbers (Scrubbers)
Absorption is the most widely used recovery tech-
nique for separation  and purification  of inorganic
vapor emissions. The removal efficiency achiev-
able with absorbers can be greater than 99 percent.
It will typically be determined by the actual concen-
trations of the specific HAP  in  gas and liquid
streams and the corresponding equilibrium con-
centrations. Table 3-2 summarizes the reported ef-
ficiencies for various inorganic vapors employing
absorption as the control method.

                      28
As discussed in Section 3.2.1.6 for organic vapors,
the suitability of absorption for controlling inorgan-
ic vapors in gaseous emission streams  is  depen-
dent on several factors. The most important factor
is the solubility of the pollutant vapor in the sol-
vent. The ideal solvent should be nonvolatile, non-
corrosive, nonflammable, nontoxic, chemically sta-
ble,  readily available,  and  inexpensive. Typical
solvents used by industry for inorganic vapor con-
trol  include water, sodium  hydroxide solutions,
amyl alcohol, ethanolamine, weak acid solutions,
and  hypochlorite solutions.  Other factors  which
may affect  inorganic vapor absorption are  similar
to those for organic vapor absorption (see Section
3.2.1.6).

Water is the ideal solvent for inorganic vapor con-
trol  by  absorption. It offers  distinct advantages
over other  solvents, the  main one being  its low
cost. It  is typically used on a once-through basis
and  then discharged to  a wastewater treatment
system. The effluent may require pH adjustment to
precipitate  metals and other HAP's as hydroxides
or salts; these are typically less toxic and  can be
more easily disposed of.


3.2.2.2 Adsorbers
When the removal of inorganic vapors is especially
difficult using absorption methods, adsorption may
prove to be more effective. Adsorbents such  as
activated carbon,  impregnated activated carbon,
silica gel, and activated alumina are capable of ad-
sorbing various  inorganic vapors  and gases. The
degree of adsorption is dependent not only on the
waste stream characteristics, but also on the differ-
ent characteristics of the adsorbents.

Carbon  adsorption, using conventional and chemi-
cally impregnated carbons, is widely used for con-
trolling  inorganic vapors such as  mercury, nickel
carbonyl,  phosgene, and amines.  For  example,
when mercury vapors are passed through a bed of
sulfur-impregnated carbon,  the mercury  vapors
react with the sulfur to form a stable mercuric sul-
fide. Over 95 percent of the mercury removed in
this way can be recovered for reuse.

Important factors to consider when  choosing an
adsorbent  for inorganic  vapor control  are very
similar  to those for organic vapor control, which
are discussed in Section  3.2.1.5.  Some of these
factors  include the amount of adsorbent needed,
temperature rise of the gas stream due to adsorp-
tion, ease of regeneration, and the  useful life of the
adsorbent.  Most of the reported removal efficien-
cies  for inorganic vapors are for activated  carbon
and  impregnated activated carbon, and range from
90 to 100 percent. Table 3-2 summarizes removal
efficiencies reported for various inorganic  vapors
controlled by adsorption.

-------
Activated carbons are the most widely used adsor-
bents for inorganic vapor control. In several cases,
they must be treated (i.e., impregnated with chemi-
cals) for effective application.  Since activated car-
bons are relatively sensitive to emission stream
conditions, pretreatment of the emission stream
may be necessary. Pretreatment methods such as
filtration, cooling, and dehumidification may be re-
quired depending on the emission stream condi-
tions. Filtration is used to prevent plugging of the
adsorber bed by any solids or  particles which may
be in the emission stream. Ideal adsorption condi-
tions for impregnated activated carbons are rela-
tive humidities less than 50 percent and gas stream
temperatures below 130°F.  Inorganic vapor con-
centrations are not recommended to exceed 1,000
ppmv (preferably, less than  500 ppmv) when acti-
vated carbon is used as an adsorbent.
3.2.3  Control Techniques for Organic/Inorganic
      Vapor Emissions From Process Fugitive
      Sources
Process fugitive emissions are defined in this hand-
book as emissions  from a process or  piece  of
equipment that are being emitted at locations other
than the main vent or process stack. Process fugi-
tive emissions include fumes  or gases which es-
cape from or through access ports  and feed and/or
discharge openings to a process. Examples include
the open top of a vapor degreaser, the slag or metal
tap opening on a blast furnace, and the feed chute
on a  ball mill. Process  fugitive emission sources
can also include vent fans from rooms or enclo-
sures containing an emissions  source. An example
would be a vent fan on a perchloroethylene dry
cleaner or the vent  fan on a  press room. Other
examples of process fugitive sources include cool-
ing towers and process drains.
                    These sources can be controlled by add-on control
                    devices once the emissions from  the sources are
                    captured  by hooding, enclosures, or closed vent
                    systems and then transferred to a control device.
                    Because of the nature of the opening (e.g., for ac-
                    cess or maintenance), the opening through which
                    emissions escape cannot  be totally  enclosed  or
                    blanked off. Operators have to  access the equip-
                    ment or materials  have to be fed or discharged
                    from the process. For this reason, hoods or partial
                    enclosures are used to control emissions from such
                    openings.

                    Proper hood design  requires a sufficient knowl-
                    edge of the process or operation so that the most
                    effective hood or enclosure can be  installed to pro-
                    vide minimum exhaust volumes for effective conta-
                    minant control. In  theory, hood design is based
                    upon  trying to enclose the process and  keep  all
                    openings to a minimum and located away from the
                    natural path of containment travel. Where possible,
                    inspection and maintenance openings should  be
                    provided  with  doors.  In practice, hoods  are de-
                    signed using the capture velocity  principle which
                    involves creation of an air flow past the source of
                    containment sufficient to remove  the highly con-
                    taminated  air from around the  source or issuing
                    from that source and  draw the air  into an  exhaust
                    hood. The capture velocity principle is based on the
                    fact that small dust particles travel very short dis-
                    tances (on the order of inches) when thrown  or
                    emitted from a source and therefore  can be as-
                    sumed to  follow air  currents. Vapors and gases
                    exhibit the same effects.
                    In practice, hood capture efficiency is very difficult
                    to  determine  and  therefore,  when  evaluating  a
                    hood,  one of the few parameters that can be con-
                    sidered is the  capture velocity. Standard design
                    values of capture  velocity are available from the
                    American Conference of Government Industrial Hy-
Table 3-2.   Current Control Methods for Various Inorganic Vapors (1)

                                       Absorption
Calcium Fluoride (CaF2)
Silicon Tetrafluoride (SiF4)
Hydrogen Fluoride (HF)
Hydrogen Bromide (HBr)
Titanium Tetrachloride (TiCI4)
Chlorine (CI2)
Hydrogen Cyanide (HCN)
                                              Adsorption
Inorganic Vapor
Mercury (Hg)
Hydrogen Chloride (HCI)
Hydrogen Sulfide (H2S)
Reported
Removal
Efficiency (%)
95
95
98
Solvent
Brine/Hypochlorite
solution
Water
Sodium carbonate/Water
Reported
Removal
Efficiency (%)
90
100
Adsorbent
Sulfur-impregnated
activated carbon
Ammonia-impregnated
 95
 95
85-95
99.95
 99
 90
Water
Water
Water
Water
Water
Alkali solution
99
                                                                              activated carbon
Calcined alumina
                                                 Ammonia-impregnated
                                                 activated carbon
                                                                       29

-------
gienists in the Industrial Ventilation Manual (see
Table 3-3).(2) Once a capture velocity has been de-
termined, the volume of air required should be
based  on maintaining this capture velocity at the
emissions point furthest from the hood. This cap-
ture velocity should be sufficient to overcome any
opposing air currents. (For additional information
on  hood design guidelines for several industries,
see Appendix A.7, reference 3, or reference 2.)

Very few measurements of hood capture efficiency
have been conducted.(4,5,6) Hood capture efficien-
cies of between 90 and 100 percent are possible
depending on the situation and the particular pro-
cess fugitive sources  being controlled. For sources
where operator access is not  needed and where
inspection doors can be provided, efficiencies to-
ward the upper end of the range are achievable. For
sources where emissions are more diffused, for
example, from printing  presses,  capture efficien-
cies of 90 percent may be difficult.  In the flexible
vinyl and printing industry, 90 percent is typically
the upper bound for  capture efficiency on coating
presses.(4) In the publication rotogravure industry,
capture efficiencies of 93 to 97 percent have  been
demonstrated based on material balances.(B)

Once the  process fugitive emissions are captured,
the selection of the control device will be depen-
dent on the emission stream characteristics, HAP
characteristics, and  the required overall perfor-
mance levels  (e.g., removal  efficiency). Note that
the required performance level for the control de-
vice will be determined  by the capture efficiency.
The factors that affect the control device  selection
process are the same as for point sources; there-
fore, refer to Sections 3.2.1 and 3.2.2.
     For process fugitive emission sources such as pro-
     cess drains, the control alternatives involve a clo-
     sure or a seal. A common method involves the use
     of a P-leg in the drain line with a water seal. A less
     common, but  more effective method,  is a com-
     pletely closed drain system. Several factors affect
     the performance of water-sealed drains in reducing
     organic emissions: drainage rate, composition and
     temperature of the liquid entering the drain, diame-
     ter of  the drain, and ambient atmospheric condi-
     tions. Emission reductions of 40 to 50 percent may
     be achieved with water-sealed drains.(7) In a com-
     pletely closed drain system, the system may  be
     pressured and purged to a control device to effec-
     tively capture all emissions. The control efficiency
     would then depend on the efficiency of the control
     device; 95 percent should be achievable.(7)

     Process  fugitive emissions from  cooling towers
     have not been reliably quantified due to  difficulties
     encountered in measuring them and no specific
     control guidelines have been developed.(8) Prob-
     ably the best control technique currently available
     is close monitoring of heat exchangers and other
     equipment to  detect small  leaks as they occur.

     3.2.4 Control Techniques for Organic/Inorganic
           Vapor Emissions from Area Fugitive Sources
     The control measures that can be employed  for
     controlling  organic or  inorganic vapor  emissions
     from area fugitive sources are basically the same.
     Organic fugitive emissions have been more exten-
     sively studied than inorganic fugitive  emissions.
     Therefore, the following discussion will  be primar-
     ily based on organic vapors. However, control tech-
     niques for inorganic vapor emissions will also be
     discussed.
Table 3-3. Range of Capture Velocities (2)
Condition of
Contaminant Dispersion
Released with practically no velocity into quiet air
Released at low velocity into moderately still air
Examples
Evaporation from tanks; degreasmg, etc.
Spray booths; intermittent container filling;
Capture
Velocity (fpma)
50-100
100-200
Active generation into zone of rapid air motion
Released at high initial velocity into zone of very rapid
  air motion
  low speed conveyor transfers; welding;
  plating; pickling
Spray painting in shallow booths; barrel fill-
  ing; booths; barrel filling; conveyor load-
  ing; crushers
Grinding; abrasive blasting; tumbling
                                                                                          200-500
500-2,000
In each category above, a range of capture velocity is shown
Lower End of Range:
- Room air currents minimal or favorable to capture.
- Contaminants of low toxicity or of nuisance value only.
- Intermittent, low production.
- Large hood-large air mass in motion.
Upper End of Range:
- Disturbing room air currents.
- Contaminants of high toxicity.
- High production, heavy use.
- Small hood-local control only.
The proper choice of values depends on several factors:
afpm = feet per minute.
                       30

-------
 Fugitive emissions of organic vapors occur in
 plants processing organic liquids and gases, such
 as  petroleum refineries,  chemical  plants,  and
 plants producing chemically based products such
 as plastics, dyes, and drugs. One group of emission
 sources found in plants of this type is  commonly
 referred to as equipment leaks. Fugitive emissions
 of this type  result from  incomplete  sealing of
 equipment at the point of interface of process fluid
 with  the environment. Control  techniques for
 equipment leaks include leak detection  and repair
 programs and equipment installation or configura-
 tion.  The following sections contain information
 about control techniques for common types of pro-
 cessing equipment found in  plants processing or-
 ganic materials. Control techniques and control ef-
 ficiencies for  common  types  of  processing
 equipment are summarized in Table 3-4.

 Table 3-4. Summary of Control Effectiveness for Controlling
        Organic Area Fugitive Emission Sources (9)
                    Control Technique    Control
 Emission Source           Equipment    Effectiveness (8)
 	Modification	(%)	
 Pumps             Monthly leak detection
                  and repair              61
                 Sealless pumps          100
                  Dual mechanical seals     100
                 Closed vent system8       100
 Valves
 - Gas             Monthly leak detection
                  and repair              73
                  Diaphragm valves        100
 - Light liquid        Monthly leak detection
                  and repair             46
                 Diaphragm valves        100
 Pressure Relief      Rupture disk            100
 Valves           Closed vent system3       100
 Open-ended Lines    Caps, plugs, blinds        100
 Compressors         Mechanical seals with
                  vented degassing
                  reservoirs             100
                 Closed vent system8       100
 Sampling Connections  Closed purge sampling     100
 "Closed vent systems are used to collect and transfer the fugitive
 emissions to add-on control devices such as flares, incinerators, or
 vapor recovery systems.

Pumps—Several types of equipment or equipment
configurations can be used to eliminate  or capture
all organic vapors leaking from pump seals.  There
are, first of all, leakless  pumps, pumps designed
with no interface between the process fluid and the
environment, such as diaphragm seal, and canned
pumps. These  pumps effectively limit fugitive
emissions.  However, they  are  limited  in applica-
tion.

Sophisticated pump seals can also be used to cap-
ture or eliminate fugitive emissions. Dual seal sys-
tems with pressurized barrier fluids or low pressure
systems vented to control devices  may be used in
some applications. Another approach which may
be used  involves venting the entire seal area to a
control device. Capture efficiencies should be virtu-
ally 100 percent in both systems vented to control
devices.  Then the overall control efficiency would
be limited by the efficiency of the control device.

Another approach to reducing (but not eliminating)
organic fugitive emissions from pumps is leak de-
tection and  repair programs. A leak detection and
repair system modeled  after the one EPA devel-
oped for the New Source  Performance Standards
(NSPS) of the synthetic organic chemical industry
(SOCMI)  should achieve about 60 percent control
efficiency.(9) The efficiency of leak detection and
repair programs is dependent on  several factors
such  as frequency of monitoring,  effectiveness of
maintenance, action level, and  underlying tenden-
cy to  leak. These factors and their effect on control
efficiency have been studied and discussed in re-
ferences  9,  10, 11  and 12. Also, models  are avail-
able for calculating the effectiveness of leak detec-
tion and  repair programs (e.g., see reference 13).

Valves—As  with pumps, control of fugitive emis-
sions from valves may be accomplished by install-
ing equipment designed to isolate the process fluid
from  the environment. But also as with pumps,
leakless valves such as diaphragm valves are limit-
ed in  their application.

Leak  detection  and repair programs have  been
used  to reduce fugitive emissions from valves. As
indicated above for pumps, the control efficiency of
a leak detection and repair program depends on
the frequency of monitoring, the effectiveness of
maintenance, the action  level, the underlying ten-
dency to leak, and other factors. A leak detection
and repair program modeled after the one devel-
oped by  EPA for the NSPS of the  SOCMI should
achieve a control efficiency of about 70 percent for
valves in gas service  and about  50  percent for
valves in light liquid service.(9)

Pressure Relief Valves—Fugitive emissions from
pressure relief valves may be virtually eliminated
through the use of rupture disks to prevent leakage
through  the seal. Fugitive emissions  may also be
added to gases collected  in a flare system by piping
the relief valve to a flare header. The control effi-
ciency, then, depends on the destruction  efficiency
of the flare. If flares are operated in accordance
with  flare  requirements  recently  established  by
EPA for sources complying with NSPS, at least 98
percent control efficiency should be achieved.

Open-ended Lines—Leakage of organic  vapors
through valve seats to  the open ends of pipes can
be eliminated by the installation of caps, plugs, or
blind  flanges. The control efficiency should be 100
percent as  long as the plugs and caps remain in
place.
                      31

-------
Compressors—Fugitive emissions from compres-
sor seals can be controlled by venting the seal area
to a flare or other control device. Barrier fluid sys-
tems can also  be used to purge the seal area and
convey leakage to a control device. Capture effi-
ciency should  be 100 percent and, therefore, the
overall control efficiency would depend on the effi-
ciency of the control device.

Sampling  Connections—Fugitive emissions from
sampling connections can be controlled by return-
ing the purged material to the process or by dispos-
ing of it in a control device. The practice of return-
ing purged material to the process in a closed
system should achieve almost 100 percent control
efficiency. The control efficiency achieved by di-
verting the collected purge material to a control
device depends on the efficiency of the device.

Other area fugitive emission sources include la-
goons and ponds where liquid waste streams con-
taining organic compounds are disposed of. Emis-
sions and emission rates of organic vapors from
such sources are not well documented, and such
sources are not easily controlled. The best method
currently available for reducing emissions from la-
goons and ponds is enhancing upstream treatment
processes,  thereby minimizing the amount of or-
ganic material reaching the  lagoons and/or ponds.

Area fugitive emissions of inorganic vapors may be
found in plants  processing inorganic  chemicals,
metals,  electronics, and other products. Although
extensive work  has not been  done  to  quantify
equipment leaks in  plants processing  inorganic
chemicals, it is expected that they would be similar
to equipment leaks encountered in plants process-
ing organic chemicals. Therefore, plants  process-
ing highly  volatile compounds  such as hydrogen
chloride or ammonia would be expected to benefit
by the same control techniques applied to reduce
or eliminate fugitive emissions containing organic
compounds. There are a couple of differences to
keep in mind, however. First, for those control tech-
niques that employ a control device to treat collect-
ed vapors, the control device will probably differ.
Instead of a combustion device, an absorber, con-
denser, or an adsorber may be a more appropriate
choice.  Another  difference would be relevant to
leak detection and repair programs. Leak detection
and  repair programs for organic vapors were de-
veloped using portable organic analyzers. Portable
analyzers that respond to the inorganic vapors of
concern would have to be used for leak detection of
inorganic materials.

Controlling inorganic vapor emissions from area
sources such as lagoons and/or ponds where liquid
waste streams containing volatile inorganic com-
pounds are disposed of is quite difficult. The best
control  method  currently available is minimizing
the quantity of inorganic compounds reaching the
lagoon and/or pond by improving upstream treat-
ment processes.

3.2.5 Control Device Selection for a Hypothetical
     Facility
This subsection illustrates the control device selec-
tion process discussed in the previous sections for
a  hypothetical  facility with several  emission
streams. Assume that the owner/operator of this
facility has requested assistance regarding the con-
trol of  these emission streams.  The data supplied
by the  owner/operator are presented in Figures 3-2
through 3-7.

Emission Stream 1 (see Figure 3-2)—This stream is
the same as that  described  in Chapter 2 in the
Example Case. Assuming the HAP control require-
ment is 99 percent reduction, from Figure 3-1, the
only applicable control technique for this level of
performance at concentration levels of —100 ppmv
is thermal incineration.  The HAP concentration is
less than  25 percent of the  LEL for the HAP (see
Table B.1-1); hence, the concentration limit indicat-
ed in Table 3-1 will not be exceeded. Also, the flow
rate of Emission Stream  1  falls in the range of
application indicated for thermal incinerators in Ta-
ble 3-1.

Emission Stream 2 (see Figure 3-3)—Assume the
HAP control requirement for Emission Stream 2 is
95 percent reduction. For this level of performance,
the applicable control techniques for inlet concen-
trations of -500 ppmv are thermal incineration,
catalytic incineration, and absorption.  If either of
the incineration techniques are applied, the con-
centration limit  indicated in Table 3-1  will not be
exceeded since the HAP concentration is less than
25 percent of the LEL for the HAP (see Table B.1-1).
The flow rate of Emission  Stream 2 falls in the
range  indicated as applicable in Table 3-1 for these
control techniques. The final selection  of the con-
trol technique should also be based on design crite-
ria (Chapter 4) and costs (Chapter 5).

Emission  Stream 3 (see Figure 3-4)—Assume the
HAP control requirement for Emission Stream 3 is
98  percent reduction. In this case, the inlet HAP
concentration falls  outside the  range indicated  in
Figure 3-1; therefore, none of the control devices in
this figure are  applicable. According  to  Section
3.2.1.3, flares can  be used  to control emission
streams with high heat contents; hence, flaring can
be considered an option. Also, if a boiler or a pro-
cess heater is available on site, it can  be used to
control Emission Stream 3.

Emission Stream 4 (see Figure 3-5)—Assume the
HAP control requirement for Emission Stream 4 is
95 percent reduction. For this level of performance,
the applicable control techniques for inlet concen-
                      32

-------
 Figure 3-2.  Effluent characteristics for emission stream #1.
                                    HAP EMISSION STREAM DATA FORM*
Cot
Loc
A.
B.
C.
D.
E.
F.
G.
H.
I.
J.
K.
L.
M.
N.
O.
U.
V.
W.
-npany Glaze Chemical Company
•atinn (StmPt) 87 Octane Drive
inty) Somewhere
(State, 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Temperature (1,P,3)
Flow Rate (1,2r3)
Pressure (1,2)
Halogen/Metals (1 7)
Applicable Regulation(s)
RpquirfH r.nntrnl LPVP!
Selected Control Methods

Plant Contact Mr- J°nn
Leake
(999) 555-5024
Anpnrw Tnntar.t Mr. Efrem Johnson


ant Identification #1 /
(a) paper coating oven
(a) process point
(a) toluene
(a) orqanic vapor
(a) 73 ppmv
(a) 28.4 mm Hg 3 77°F
(a) insoluble in water
(a) provided
(a) 92 Ib/lb-mole
2% vol.
12C°F
15,000 scfm (max)
atmospheric
none / none




No.

#3
(b)
(b)
(b)
(b)
(b)
(b)
(b)
(b)
(b)
P.
Q.
R.
S
T.




. of Emission Streams Ur
Oven Fxhaust
-
-
-
-
-
-
-
-
-
Organic Content (1)***
Heat/02 Content (1)
Particulate Content (3)
Particle Mean Diam (3)
Drift Velocity/S03 (3)




iHpr Rpview


(r)
(r-)
(c)
(c)
(<•-)
(c)
(^)
(c-)
(c)
44 ppmv CH4, 4 comv ot
0.4 Btu/scf / 20.6% vol.


/




   *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
    method(s).  Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

  **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
       1 = organic vapor process emission
       2 = inorganic vapor process emission
       3 = particulate process emission

  ""Organic emission stream combustibles less HAP combustibles shown on Lines D and F.
trations of -1,000 ppmv are thermal incineration,
catalytic incineration, carbon adsorption, and ab-
sorption.  If either of the incineration methods or
carbon adsorption is  applied, the  concentration
limit indicated in Table 3-1 will not be exceeded
since the HAP concentration is less than 25 percent
of the  LEL (see Table B. 1-1). The flow rate of Emis-
sion Stream 4 falls in the range indicated as appli-
cable in Table 3-1 for these control techniques. The
final selection of the control technique should also
be based  on design criteria (Chapter 4) and costs
(Chapter 5).

Emission  Stream 5 (see Figure 3-6)—Assume the
HAP control requirement for Emission Stream 5 is
98 percent  reduction. Since this emission stream
contains inorganic HAP's,  incineration techniques
are not applicable. The only control technique that
is applicable for this level of performance and inlet
HAP concentrations of -20,000 ppmv  is absorp-
tion. The flow rate of Emission Stream 5 falls in the
range indicated as applicable for absorption.

Emission  Stream 6 (see Figure 3-7)—Assume the
HAP control requirement for Emission Stream 6 is
90  percent. Also assume that the owner/operator
has indicated his preference to recover the HAP in
the emission stream. For this level of performance,
the applicable control techniques for inlet concen-
trations of —13,000 ppmv are absorption  and con-
densation. The final  selection among these tech-
niques  should also  be  based on  design criteria
(Chapter 4) and costs (Chapter 5).

3.3 Particulate Emissions Control
Section 3.3.1  discusses add-on paniculate control
devices and  presents guidelines that  are used to
determine the applicability of each control device.
Section 3.3.2 discusses control techniques that are
used to reduce fugitive particulate emissions.

3.3.1 Control Techniques for Particulate Emissions
     from Point Sources
Three types of control devices applicable to parti-
culate-laden emission streams from point sources
are discussed below: fabric  filters (baghouses),
electrostatic precipitators (ESP), and venturi scrub-
bers. The control  efficiencies and applicability of
these devices depend on the  physical and chemi-
                                                                         33

-------
Figure 3-3.  Effluent characteristics for emission stream #2.
                                   HAP EMISSION STREAM DATA FORM*
Cor
Loc
A
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
0.
U.
V.
W.
npany Glaze Chemical
Company
atmn(StrPPt) 87 Octane Drive
(Pity) Somewhere
(State, 7ip)

Emission Stream Number/P
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Temperature (1,2,3)
Flow Rate (1,2,3)
Pressure (1,?)
Halogen/Metals (1,7)
Applicahlp Rpgnlation(s)
Rpquirpri Control I evpl
Selected Control Methods



Plant Contact M>". John
Leake
To|oph^ne M^ (999) 555-5024
Anpnrv Contact Mr- Efrem
Johnson
No nf Fmksinn Strpamc; I Inripr Rpvipvu

lant
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)
(a)









Identification #2 /
metal coating oven
process point
toluene
organic vapor
500 ppmv
28.4 mm Hg @ 77° F
insoluble in water
provided
92 Ib/lb-mole
2% vol.
120°F
20,000 scfm (max)
atmospheric
none / none

assume 95% removal
thermal incineration,


#1 Oven Exhaust
(h)
(b)
(b)
(b)
(h)
_. (b)
(h)
(h)
(b)
P. Organic Content (1)***
Q. Heat/02 Content (1)
R Partir.nlatP TontPnt (-^)
S. Particle Mean Diam. (3)
T Drift Velnfity'^n3 (i)




H

(c)
(c*
(r.\
(c)

(c)
(c)
none
2.1 Btu/scf / 20.6% vol.


1


catalytic incineration, absorption


  *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
   method(s).  Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

  **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 =organic vapor process emission
      2 = inorganic vapor process emission
      3 = paniculate process emission

 **Organic emission stream combustibles less HAP combustibles shown on Lines D and F.
cal/electrical properties of the airborne participate
matter under  consideration. Brief descriptions  of
each of these control devices appear in the subsec-
tions that follow.

Selection of  the control devices themselves de-
pends on the specific stream  characteristics and
the parameters (e.g., required collection efficiency)
that affect the applicability of each control device.
Table  3-5  identifies some  key  emission stream
characteristics that affect the applicability of each
device. Matching the characteristics of the emis-
sion stream  under consideration with the corre-
sponding information  presented in  Table 3-5 will
identify those techniques most suited to control the
stream. This does not imply that a given control
device should be excluded at this point, however.
In general, the parameters  listed in Table 3-5 are
given as typical guidelines and should not be taken
as absolute, definitive values. Gas stream pretreat-
ment equipment can be installed upstream of the
control  device  (i.e., cyclones,  precoolers,  pre-
heaters) which enables the emission stream to fall
within the parameters outlined  in Table 3-5.
The temperature of the emission stream should be
within 50 to 100°F above its dew point if the emis-
sion stream is to be treated (i.e., particulate matter
collected) by an ESP or a fabric filter. If the emis-
sion stream temperature is below this range,  con-
densation can occur; condensation can lead to cor-
rosion  of  metal surfaces, blinding  and/or
deterioration of fabric filter bags,  etc. If the emis-
sion stream is above this range, optimal  HAP col-
lection may not occur; by lowering the emission
stream temperature, the vapor component of the
HAP is reduced and, thus, an ESP or fabric filter will
collect the HAP more effectively. Procedures for
determining the dew point of an emission stream
are provided in  Appendix B.1. (For discussions of
gas stream pretreatment equipment, see reference
15 and Appendix B.11 of reference 3.)

Table 3-6 identifies general advantages and disad-
vantages for each particular control device. Table
3-6 is  used to provide additional information on
other considerations that, while not necessarily af-
fecting the technical feasibility of the control device
for the stream, may affect the overall desirability of
                       34

-------
 Figure 3-4.   Effluent characteristics for emission stream #3.
                                         HAP EMISSION STREAM DATA FORM*
Coi
Loc
A.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
0.
mpany Glaze Chemical Company
-atinn (Street) 87 Octane Drive
(Pity) Somewhere
(State 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Temperature (1,7,3)
Flow Rate (1.2.3)
Pressure (1,2)
Halogen/Metals (1 7)
Plant Contact Mr. John Leake
Toi-nh^M,, (999) 555-5024
Anenr-u Pnntart Mr. Efrem Johnson


lant Identification #1 /
(a) absorber vent
(a) process point
(a) methylene chloride
;ai organic vapor
(a) 44.000 ppmv
(a) 436 mm Hq @ 77°F
(a) insoluble in water
(a) not given
(a) 85 Ib/lb-mole
none
100°F
30,000 scfm (expected)
atmospheric
none / none
No. of Emission Streams U
nder Review

6

Arptaldphyde Manufacturi na Absorber Vent
-- (b)
(b)
(hi
(b)
(h)
(b)
(h)
(h)
(b)
P. Organic Content (1)***
Q. Heat/02 Content (1)
R. Particulate Content (3)
S. Particle Mean Diam (3)
T. Drift Velocitv/SO-, (3)
(c)

(c)
(c)
(c)
(c)

(f)
(c)
17.8% vol. C
180 Btu/scf



-
-
-
-
-
-
-
-
-
Hd
/ none


/
U.  Applicable Regulation(s)	
V.  Required Control Level —
W.  Selected Control Methods
                      assume 98% removal
                      flare, boiler, process  heater
  *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
    method(s).  Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

 **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
       1 =organic vapor process emission
       2 = inorganic vapor process emission
       3 = paniculate process emission


 '"Organic emission stream combustibles less HAP combustibles shown on  Lines D and F.
Table 3-5.    Key Characteristics for Particulate Emission Streams
Control
Device
Baghouse






ESP

Achievable
Efficiency
Range
Up to 99 + %






Up to 99 + %

Particle
Size
Limitation
Least efficient with
particles 0.1 \i.m to
0.3 (xm diameter.




Generally least
efficient with
Temperature
Dependent
on fiber type
but not
exceeding
550°F
without a
precooler
Generally up
to 1,000°F
Corrosiveness/
Resistivity
Special fiber types necessary to
resist corrosion.





Corrosion resistant materials
required. May require
Moisture
Content
Poor efficiency with
emission streams of high
moisture content, very
sensitive to changes in
moisture content of an
emission stream.

Can control streams with
relatively high moisture
Venturi
Up to 99 + %
                            particles ranging in
                            size from 0.2 jj.m to
                            0.5 (Am diameter.
Generally operates
best with particles
>0.5 |j,m diameter.
No general
limitations
conditioning agents for highly
resistive particles. Additionally,
ESP's are not used to control
organic matter since this
constitutes a fire hazard.
Special construction may be
required for corrosive emission
streams.
content (i.e., 34%vol) if so
designed, but sensitive to
moisture changes of an
emission stream.

Not sensitive to changes in
moisture content of
emissions stream.
                                                                                     35

-------
Figure 3-5.   Effluent characteristics for emission stream #4.
                                   HAP EMISSION STREAM DATA FORM*
Cor
Loc
A.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
O.
U.
V.
W.
npany Glaze Chemical Company
•atinn fStrPPt) 87 Octane Drive
(Pity) Somewhere
(Rtatp 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Tp.mppratnrp (1,2,3)
Flnw Ratfi (1,2,3)
Prpccurp (1 2)
Hq|ngpn/MRtals(1,9)
Applicable Regulation(s)
Required Control I PVP!
Selected Control Methods

Plant Contact Mr- John
Leake
T0,o.h.np Nn (999) 555-5024
Agpnr.y Contact Mr. Efrem Johnson
Nn nf Fmissinn Strpams I InHpr Rpvipuv

ant IHpntifiratinn #4 i
(a) printing press
(a) process point
(a) toluene
la^ organic vapor
la) 1,000 DDmv
(a) 28.4 mm Hg @ 77"F
(a) insoluble in water
(a) provided
(a) 92 Ib/lb-mole
40% rel. humiditv
90° F
15,000 scfm (max)
atmospheric
none / none

assump 9K rpmnval
thermal incineration,


' #1 Printing Press Exhaust
(b)
(h)
(b)
(b)
(h)
(h)
(h)
(h)
(h)
P. Organic Content (1)***
Q. Heat/02 Content (1) 4'
R Partir.iilatfi T.nntpnt (3)
S Partirle Mean Piam (3)
T Drift Velocity/S03 (3)


catalytic incineration, carbon



(r)
(0
(c)
(c)
(r.)
(^)
(r)
(<0
(c)
none
2 Btu/scf / 20.6% vol.


/


adsorption, absorption

  'The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
   method(s). Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.
  'The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 = organic vapor process emission
      2 = inorganic vapor process emission
      3 = paniculate process emission

  'Organic emission stream combustibles less HAP combustibles shown on Lines D and F.
using the device for a given emission stream. Thus,
Tables 3-5 and 3-6 used in  conjunction provide
guidelines to determine if a particular control de-
vice could and should be used for a given emission
stream. Further design criteria (Chapter 4) and cost
(Chapter 5) must be considered to  enable a com-
plete  technical evaluation of the  applicability of
these devices to an emission stream.
3.3.1.1  Fabric Filters
Fabric filters, or baghouses, are an efficient means
of separating particulate matter entrained in a gas-
eous stream. A fabric filter is typically least efficient
collecting particles in  the range of 0.1 to 0.3 (im
diameter. For particles larger and smaller than this,
a fabric filter can collect more than 99 percent of
particles in an emissions stream. Fabric filters used
to control emissions containing HAP's  have two
special constraints. First, they  should  have  a
closed, negative-pressure (suction) configuration
to prevent accidental release of the gas stream and
captured HAP's; and second, of the three principal
fabric cleaning methods (i.e., mechanical shaking,
reverse air flow, and pulse-jet cleaning), pulse-jet
type cleaning is not recommended for HAP control
situations.

As explained in Section 4.9, pulse-jet cleaned filters
are not as efficient as those cleaned by other meth-
ods, and emissions are not as constant over a filtra-
tion cycle as those from filters using the other two
cleaning methods. However, pulse-jet cleaning is
widely used  in general  industrial fabric filter appli-
cations and, therefore,  Section 4.9 does  include
information on pulse-jet cleaning. Further, Chapter
5 includes pulse-jet cost information to  allow for
review of permit applications indicating the use of
pulse-jet baghouses.

Fabric filters using mechanical  shaking or  reverse
air cleaning are fundamentally different from ESP's
and venturi scrubbers in that they are not "efficien-
cy" devices. A properly designed and operated fab-
ric filter using one of these two cleaning methods
will yield a relatively constant outlet particle con-
centration, regardless of inlet  load changes.  The
typical outlet particle concentration range is  be-
tween 0.003 to  0.01  grains/scf (gr/scf),  averaging
                       36

-------
Figure 3-6.   Effluent characteristics for emission stream #5.
                                    HAP EMISSION STREAM DATA FORM'
Cor
Loc
A.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
0.
U.
V.
W.
-npany Glaze Chemical Company
-atinn (StmPt) 87 Octane Drive
ir.ity) Somewhere
(State, 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Tpmppraturp (1,7,3)
Flow Rate (1.2.3)
Prpssnrp (1,2)
Halngpn/Mptals(1 9)
Applicable Regulation(s)
Required Control Level
Selected Control Methods


Plant Contact Mr- J°hn
Leake
Telephone No <999> 555-5024
Anpnrv Tnntar.t Mr. Efrem
Johnson
Nn of Fmissinn Strpamc I Indpr Rpuipuu

lant IHpntifir.atinn #5 /
(a) evaporator off-qas
(a) process point
(a) ammonia
(.,) inorganic vapor
(a) 20,000 ppmv
(a) 8.46 atm. @ 68°F
(a) provided
(a) not given
la) 17 Ib/lb-mole
2% vol.
85° F
3,000 scfm (max)
atmospheric
none / none

assume 98% removal
absorption



Urea Fvapnratnr Off-ga^ Fxhaust.
(h)
(h)
(b)
(b)
(h)
(b)
(h)
(h)
(b)
P. Organic Content (1)***

R. Particulate Content (3)
S Particle Mean Diam (3)
T Drift Ve|nrity/Rn3 (3)




(r)

(c)
(c)

(p)
(r)
(^)
(c)

/


/




  *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
   method(s). Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

 **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 = organic vapor process emission
      2 = inorganic vapor process emission
      3 = paniculate process emission

 ** Organic emission stream combustibles less HAP combustibles shown on Lines D and F.
around 0.005 grains/scf. These numbers  can be
used to ascertain an expected performance level
(see the following Example Case). This is not meant
to be an absolute, definitive performance level. A
vendor should assist in any attempt to quantify an
actual performance level.
Variables important  to achieve a given  perfor-
mance (i.e., air-to-cloth ratio, cleaning mechanism,
fabric type) are discussed in detail in Section 4.9.
Fabric filters are sensitive to emission stream tem-
perature and a precooler or preheater may be re-
quired, as discussed previously.  Fabric filters oper-
ate  at low  pressure  drops,  giving them  low
operating costs. They are generally not a feasible
choice to control emission  streams  with  a high
moisture content.

3.3.1.2 Electrostatic Precipitators
Electrostatic precipitator particle removal occurs by
charging the particles, collecting the particles, and
transporting the collected particles into a hopper.
ESP's are  less sensitive to particle  size than the
other two devices, but very sensitive to those fac-
tors that affect the maximum electrical power (volt-
age) at which they operate. These are principally
the aerosol density (grains/scf)  and the electrical
resistivity of the material. The electrical resistivity
of the particles  influences the drift velocity, or the
attraction between the particles  and the collecting
plate. A high resistivity will cause a low drift veloc-
ity which will decrease the overall collection effi-
ciency. Electrostatic precipitators are discussed fur-
ther in Section 4.10.

3.3.1.3 Venturi  Scrubbers
Venturi scrubbers use an  aqueous  stream to re-
move particulate matter from an  emissions stream.
The performance of a venturi scrubber is not affect-
ed by sticky, flammable, or corrosive particles. Ven-
turi scrubbers are more sensitive to particle size
distribution than either ESP or fabric filters. In gen-
eral, venturi scrubbers perform most efficiently for
particles above 0.5 |xm in diameter (see  Section
4.11 for  further detail). Venturi  scrubbers have a
lower initial cost than either fabric filters or ESP's,
but the high pressure drop required for high collec-
tion efficiencies  contributes to  high operating
costs.
                                                                           37

-------
                                          Example Case
  Assume a facility is required to achieve an emis-
  sion limit for particulate emissions from a mu-
  nicipal waste incinerator. The emissions stream
  particles  consist primarily of fly ash; however,
  10 percent of these particles is a HAP: cadmium.
  The characteristics of the emission stream after
  exiting a  heat exchanger are shown in Figure 3-
  8. From Figure B.1-1 (Appendix B), the dew point
  of an emission stream containing 200 ppmv S03
  and 5 percent moisture is approximately 327°F.
  Thus, the emission  stream temperature (400°F)
  is within  50 to 100°F above its dew point, thus
  minimizing the amount of the HAP in vapor form
  and eliminating condensation problems.
  Calculate the allowable outlet particle  concen-
  tration:
              0.005 gr v  110,000 acf
              	      	
                                                                             530°R-scf
              x
   scf
   60 min-lb
                             mm
                                                                         (400  + 460)°R-acf
                                                                  I b
                                                                  hr
                7,000 gr-hr

              To calculate the HAP outlet emission rate:
                                               2.9   -x 0.10 = 0.29
                                                    hr
                                   lbHAP
                                     hr
                             Ib
110,000 acf    3.2 gr    	
   min        acf     7,000 gr


- 3'017 HF
60 min
  hr
   3,017 -j^x (1 - 0.999) = 3.017 -]£-

  Since the HAP constitutes 10 percent of the total
  particulate matter, the outlet  concentration of
  the HAP is:
   0.10 x 3.017    -= 0.3017
                            lb
  This value assumes that the HAP is in particulate
  form and that it is collected as efficiently as the
  other particles.

  A fabric filter will  generally control particles to a
  limit of 0.005 gr/scf. Converting gr/scf to  Ib/hr
  yields:
This value again assumes that the HAP is in
particulate form and that it is collected as effi-
ciently as the other gas stream particles. The
calculated HAP emission rate  from a fabric filter
is  less than the allowable rate, indicating that
use of a fabric filter is an appropriate control
technique for this emission stream.

In  general,  an  electrostatic  precipitator can
achieve control efficiencies of 99.9 percent, pro-
vided the particle resistivity is  not "high." The
drift velocity of the particles (0.30 ft/sec) is indic-
ative of particles with an "average" resistivity;
therefore, an ESP can probably be used to con-
trol this stream and, thus, it also is an appropri-
ate control technique for this emission stream.

A venturi scrubber has difficulty controlling par-
ticles below 0.5 |xm diameter. The emission
stream  presented contains particles generally
above this limit; therefore, a venturi scrubber
may be  an  acceptable control device for this
example and, thus, all three  particulate control
devices are  appropriate control techniques  for
this emission stream.  To determine the basic
design parameters and actual applicability of
each control device, Section  4.9 (Fabric Filter),
Section 4.10 (ESP's), and Section 4.11 (Venturi
Scrubbers) must be examined.
3.3.2 Control Techniques for Particulate Emissions
     from Fugitive Sources
Fugitive emission  sources may be broken down
into two source categories: process sources and
area sources, as defined in Section 2.2. The meth-
ods used to control process sources of fugitive par-
ticulate emissions are generally  different from
those applied to area  sources. Basically, process
fugitive sources can  employ  conventional  mea-
sures (i.e., capture techniques  and add-on control
devices) while area fugitive sources either cannot
use conventional measures or the  use of conven-
tional measures is precluded due to cost. For exam-
ple, fugitive emissions from unpaved roads cannot
use conventional  control measures, but fugitive
emissions from  area sources such as pumps and
valves can be captured  and  ducted to  a control
device, although the costs may be prohibitive. Area
             sources  are  often controlled  by preventive tech-
             niques rather than capture/control techniques.

             Section  3.2.3 discusses methods of hooding  and
             capture of process emissions. The remaining fugi-
             tive particulate emission  control methodologies
             (i.e., nonconventional  techniques) can be applied
             to  multiple fugitive emission sources—both for
             process and area sources.

             The following subsections discuss the different
             types of fugitive particulate emission control tech-
             niques that can be applied to  general process  and
             area  fugitive particulate  emission  sources (i.e.,
             sources common to many industries). (See Appen-
             dix A.8,  reference  3, for  industry-specific  and
             source-specific information for fugitive particulate
             emission control and Appendix A.9, reference 3, for
             information on chemical dust suppressants.)
                      38

-------
Figure 3-7.  Effluent characteristics for emission stream #6.
                                   HAP EMISSION STREAM DATA FORM*
Cor
Loc
A
<"-\.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L.
M.
N.
0.
npany Glaze Chemical Company
:atmn (Street) 87 Octane Drive
irityl Somewhere
(State, 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Moisture Content (1,2,3)
Temperature (1,7, 3)
Flow Rate (1,2,3)
Pressure (1.2)
Halogen/Metals (1 7)
Plant Contact Mr- John
Leake
(999) 555-5024
Agenry Cnntar.t Mr. Efrem Johnson
Nn nf Fmkcinn Streams I Inrier Review;

lant
(a)
(a)
la)
(a)
(a)
(a)
(a)
(a)
(a)





Identification #6 /
condenser vent
process point
styrene
organic vapor
13,000 ppmv
provided
insoluble in water
not given
104 Ib/lb-mole
negl igible
90°F
2,000 scfm (max)
atmospheric
none / none


Styrene Recovery Condenser Vent
(b)
(h)
(hi
(b)
(h)
(h)
(h)
(h)
(h)
P. Organic Content (1)***
Q. Heat/02 Content (1) 61>
R. Particulate Content (3)
s Partible Mean Diem (3)
T. Drift Velocitv/SOn (3)
If)
(<••)
(r\
(c)
-(f-)
- (c)
(f)
(r)
(c)
none
5 Btu/scf / 20.7% vol.


/
U.  Applicable Regulation(s)	
V.  Required Control Level	
W.  Selected Control Methods
assume 90% removal
absorption,  condensation
  *The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
   method(s) Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

 **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 = organic vapor process emission
      2 = inorganic vapor process emission
      3 = paniculate process emission

 **0rganic emission stream combustibles less HAP combustibles shown on Lines D and F.
An extensive review of available literature on fugi-
tive emissions revealed that one reference included
almost all  necessary information  pertinent to  the
scope of this handbook. Consequently, most of the
following subsections are taken directly from  the
following document: Technical Guidance for Con-
trol of Industrial Process Fugitive Particulate Emis-
s/"ons.(16) An April 1985 draft final document from
EPA on fugitive emissions (17) has been distributed
and was  reviewed. Although  no  significant
changes were made to this  section as a  result of
that document, it is referenced and does provide
the reader with another comprehensive view of the
subject.


Throughout the  discussions,  control efficiencies
are stated for many  of the control techniques. It is
important to note that the efficiency values are esti-
mates. The ability to quantify accurately the emis-
sion rates from a fugitive emission source has  not
yet been fully realized.
                   3.3.2.1 Process Fugitive Particulate Emission
                          Control
                   Control of HAP process fugitive emissions may be
                   accomplished by capturing the paniculate material
                   and venting it to an  add-on control  device  (i.e.,
                   venturi scrubbers, fabric filters, and ESP's). Venting
                   emissions is accomplished by exhausting particle-
                   laden air through fixed or movable  ducting under
                   negative  pressure. The airflow into  the  ducting
                   must be sufficient to maintain particle size capture
                   velocity and to overcome opposing air currents.
                   The effect of opposing air currents can be eliminat-
                   ed  by complete enclosure, or  can be reduced by
                   minimizing  the opening of capture  enclosures, or
                   by utilizing  curtains or partitions to block room air
                   currents. If enclosure or installation of fixed hoods
                   is not feasible due to space limitations or oper-
                   ational procedures, movable hoods may be a  via-
                   ble alternative. Movable hoods can be placed over
                   the fugitive emissions source  as the  production
                   cycle permits. For example, movable hoods can be
                                                                         39

-------
Table 3-6.    Advantages and Disadvantages of Particulate Control Devices

                              Advantages
                      Disadvantages
Baghouse    —Very efficient at removing fine participate matter from
             a gaseous stream; control efficiency can exceed 99%
             for most applications.
            —Lower pressure drop than venturi scrubber when
             controlling fine particulates; i.e., 2" to 6" H20 compared
             with a40" H20.
            —Can collect electrically resistive particles.
            —With mechanical shaking or reverse air cleaning,
             control efficiency is generally independent of inlet
             loading.
            —Simple to operate.
ESP         —Can control very small (<0.1 (im) particles with high
              efficiency.
            —Low operating costs with very low pressure drop
              (0.5" H20).
            —Can collect corrosive or tar mists.
            —Power requirements for continuous operation are low.
            —Wet ESP's can collect gaseous pollutants.
Venturi      —Low initial investment.
            —Takes up relatively little space.
            —Can control sticky, flammable, or corrosive matter with
              few problems.
            —Can simultaneously collect particulates and gaseous
              matter.
            —Control efficiency is independent of particle resistivity.
            —Simple to operate, few moving parts.
    —Cannot control high temp stream (>550°F) without a
      precooler.
    —Cannot effectively control stream with high moisture
      content.
    —Highly erosive particles can damage the filter.
    —Mechanical collectors generally required upstream if
      significant amounts of large particulates (>20 n-m) are
      present.
    —Needs special or selected fabrics to control corrosive
      streams.
    —Least efficient with particles between 0.1 n.m to 0.3 |xm
      diameter.

    —High initial capital investment.
    —Not readily adaptable to changing conditions.
    —Conditioning agents may be necessary to control
      resistive particles.
    —More sensitive to particle loading than other two
      devices
    —Space requirements may be greater than that for a
      fabric filter or venturi scrubber.

    —High operating cost due to high pressure drop. (40" H2O
      or greater), particularly for smaller (<1 n-m) particles.
    —Has wastewater and cleaning/disposal costs.
    —Least efficient with particles less than 0.5 jim diameter.
placed over the filling hatch in some types of trucks
and rail cars during  loading. Movable hoods have
also been applied in the production of coke, where-
by  a movable  hood follows quench cars during
coke pushing. Another alternative is to evacuate an
enclosed building to a control device.
Once the emissions  are captured, the selection of
the control device will be dependent on the emis-
sion  stream  characteristics,  HAP characteristics,
and the required performance levels (i.e., removal
efficiency). It is important  to  remember that  the
required performance level for the control device is
influenced by the efficiency of the capture system.
The factors that affect the control device selection
process are the same as for point sources, and are
discussed in Section 3.3.1.

3.3.2.2  Area Fugitive Emission Control From
        Transfer and Conveying
Loss of material from conveyors is primarily at the
feeding, transfer, and discharge points and occurs
due to spillage or windage. The majority of particu-
late emissions are generally from spillage and me-
chanical  agitation of the material  at  uncovered
transfer points. However,  emissions  from  inad-
equately enclosed systems can be quite extensive.
Table 3-7 presents control techniques applicable to
these emission sources.

Control by wet suppression methods includes the
application  of  water, chemicals, and  foam. The
point of application is most commonly at the con-
veyor feed and discharge points, with some appli-
cations at conveyor transfer points.  Wet suppres-
sion  with water  only  is  a relatively inexpensive
technique; however, it has the inherent disadvan-
tage  of being short-lived. Control with chemicals
(added to water for improved wetting) or foam is
longer lasting but  more expensive than water
alone.

Foam is  effective  in  dust  suppression  because
small particles (in the range of 1 to 50 |xm diameter)
break the surface of the bubbles in the foam  when
they come in contact, thereby wetting the particles.
Particles  larger than 50 |xm only move the bubbles
away. The  small wetted  particles then  must  be
brought together or brought in contact with  larger
particles  to  achieve agglomeration.  If foam  is  in-
jected into free-falling aggregate at a transfer  point,
the mechanical motion provides the required parti-
cle to bubble contact  and subsequent particle-to-
particle contact.

Highly diluted chemical wetting agents are applied
by water  jets ahead of any points in the conveying
system where  dusting occurs. The wetting  agent
breaks down the surface tension of  the water,  al-
lowing it to  spread further, penetrate deeper, and
wet the small particles better than untreated water.
With  mechanical agitation of the material, the small
particles  agglomerate. For  effective control,  the
                        40

-------
Figure 3-8.   Effluent characteristics for a municipal incinerator emission stream.

                                   HAP EMISSION STREAM DATA FORM*
Cor
Loc
A.
B.
C.
D.
E.
F.
G.
H.
1.
J.
K.
L
M.
N.
0.
-npany Incineration, Inc.
•af,nn'(StrPPt! 123 Main Street
irityl Somewhere
(Rtatp, 7ip)

Emission Stream Number/Pi
HAP Emission Source
Source Classification
Emission Stream HAP's
HAP Class and Form
HAP Content (1,2,3)**
HAP Vapor Pressure (1,2)
HAP Solubility (1,2)
HAP Adsorptive Prop. (1,2)
HAP Molecular Weight (1,2)
Mnisttirp Content (1 7 3)
Temperature (1,? 3)
Flow Rate (1.2.3)
Prpssum (1,2)
Haloqen/Metals(1,2)
Plant Cnntar.t Mr. Phil
Brothers

TplPnhoneNo. (999) 555-5624
Anonrv, Pnntart Mr. Ben Hold


No. of Emission Streams Ur
iripr Rpvifiw

1

(ant |Hpntifi<-ptinn #1 / Incineration Exhaust
(a) municipal incinerator
(a) process point
(al cadmium
/ 1 inorganic particulate
(a) 10%
(a)
(a)
(a)
(a)
5% vol.
400° F
110,000 scfm
atmospheric
none / none
(b)

(b)
(b)

(hj
(h)
(h)
(h)
P. Organic Content (1)***
Q. Heat/02 Content (1)
R. Particulate Content (3)
S Particle Mean Diam (3).
T. Drift Velocity/So, (3)
(r)

(c)
(c)
(r.)
(c.)

(r)
(0


3.2 qrains/acf.
1.0 urn
0.30 ft/sec /
-
-
-
-
_
-
-
-
-


flvash

200 ppmv
u.
V.  Required Control Level	
W.  Selected Control Methods .
                                assume 99.9% removal
                                fabric filter, electrostatic precipitator, venturi scrubber
  •The data presented are for an emission stream (single or combined streams) prior to entry into the selected control
   method(s).  Use extra forms if additional space is necessary (e.g., more than three HAP's). and note this need.

  **The numbers in parentheses denote what data should be supplied depending on the data in steps "C" and "E":
      1 = organic vapor process emission
      2 = inorganic vapor process emission
      3 = particulate process emission

 ***0rgamc emission stream combustibles less HAP combustibles shown on Lines D and F.
spray should  be applied  at each  point where the
particles might be fractured, allowed to free fall, or
subjected to strong air currents.


3.3.2.3 Area Fugitive Emission Control From
       Loading and Unloading
Loading and unloading bulk material is common to
many processing  industries. Loading  and unload-
ing operations can be either for external transpor-
tation of material to or from a facility or for internal
transportation within a facility (for example, inter-
nal transportation might consist  of  loading of a
mining haul truck with ore via a front-end loader for
subsequent unloading to  a crushing process). (See
reference 16  or Appendix A.8, reference  3, for in-
dustry-specific information on loading and unload-
ing for internal transportation.)

Various control technology applications for loading
and unloading operations are presented in Table 3-
8. These techniques can be used alone or at times
in  various combinations. Generally, the simulta-
neous use of more than one technique will provide
increased levels of control.
                                                   Rail car and truck loading —To minimize particulate
                                                   emissions from rail car and truck loading, the entire
                                                   operation  can be enclosed by the use of doors on
                                                   the loading shed. This prevents a wind tunnel effect
                                                   and allows dust emitted in the enclosure to settle to
                                                   the ground within the  enclosure. By venting the
                                                   entire enclosure to  a control device, dust leakage
                                                   around the doors and any other openings can be
                                                   prevented, thus ensuring near 100 percent control.

                                                   Exhausting the car or truck body to a dust removal
                                                   device reduces emissions if the body is fairly well
                                                   enclosed. In open type rail or truck bodies this tech-
                                                   nique is not too effective.

                                                   Choke-feed eliminates free fall of material into the
                                                   car or truck. In this technique the mouth of the feed
                                                   tube  is immersed in the material being unloaded.
                                                   This  technique only works for fairly free-flowing
                                                   dry material.  A telescopic chute or spout also es-
                                                   sentially eliminates the free-fall distance of the ma-
                                                   terial  being  loaded. This type of system  can  be
                                                   used on all types of material. Both the choke-feed
                                                   and telescopic chute methods are only partially ef-
                                                   fective in eliminating emissions since the surface of
                                                                          41

-------
Table 3-7.   Control Technology Applications for Transfer and
          Conveying Sources (16)
Emission Points
        Control Procedure
Conveyor System
(belt, bucket
elevator, etc.)
Transfer and
Transition Points
Enclosure
1. Top covered (marginal control)
2. Sides and top covered (good
  control)
3. Completely enclosed (excellent
  control)
Wet suppression (water, chemical, foam)
  at conveyor feed points
Belt scrapers and wipers
Mechanical belt turnovers
Replacement with pneumatic system or
  screw conveyor

Enclosure
Hoods, covers, or canopies with exhaust
  to removal equipment (fabric filters,
  and wet-collectors)
Wet suppression (water, chemical, foam)
the loaded material is constantly disturbed by new
material. This surface is subject to wind and dust
entrainment.

Movable hoods exhausted to a dust removal sys-
tem can be placed over the filling hatch in some
types  of trucks and railcars during  loading.  By
keeping other openings on the body closed, any
dust generated in loading must be emitted through
the single  open hatch. A hood with sufficient air-
flow mounted around this opening could capture
most of the dust generated.

Wet suppression techniques, when applied to load-
ing operations,  can  reduce airborne dust to some
extent. The loading process naturally  breaks  up
surface coatings, but some small dust particles will
adhere to  larger pieces  so as  not to become en-
trained. Many materials cannot be readily wetted
and this technique  could  not  be used for these
materials.

Barge  and ship  loading—Due to their larger size,
barge  and ship loading  present unique problems
for dust control.  However,  a  number  of  control
techniques have been developed  and utilized,  es-
pecially at some of the larger shipping terminals.

The use of tarpaulins or similar covers over hatches
on ships and enclosed barges reduces airborne
emissions by  preventing  their escape. Air, dis-
placed by the material  being  loaded, causes the
hold to become  slightly pressurized during loading,
and the hold must be vented at some point if the
hatches are air-tight. Thus, a more effective control
system incorporates an exhaust system  for the
hold. This exhaust system is connected to a dust
control system such as fabric filter with the collect-
ed material being returned to the hold. Such a sys-
tem can practically  eliminate loading emissions if
carefully  maintained and  properly operated. The
use of a canopy hood and exhaust system over the
loading boom is less effective than a totally en-
closed system, but can still reduce emissions and is
a viable alternative for open barges. Effective utili-
zation of this technique requires some type of wind
break to increase the hood capture efficiency.

Choke feed and telescopic chutes or spouts as pre-
viously described can also be used for loading both
enclosed and open ships or barges. Wet suppres-
sion techniques  may also  help  reduce airborne
emissions if the product specifications do not pro-
hibit use of this technique.

Rail car and truck unloading — Many of the unload-
ing dust control techniques are identical to the
loading techniques. When a rail car or truck is tilted
and materials are dumped into  an  underground
chamber through a grating, exhausting air from
this chamber through a control device will effec-
tively  reduce  emissions.  By causing  air to flow
down through the grating, dust emissions are con-
tained. The face velocity of air through the  grating
is a critical design parameter in this technique. Un-
loading cars with a screw conveyor causes  less
distribution  of the material and thereby less dust.
Problems of material handling and  time require-
ments limit the application of this technique. Pneu-
matic unloading of very fine materials is an effec-
tive and widely used  technique that practically
eliminates dust emissions. With this system, care-
ful maintenance of hose fittings and the fabric filter
through which the conveying  air exhausts is re-
quired.

Barge and ship  unloading—Control of barge and
ship unloading requires enclosure of the receiving
point on the shore and  possibly exhausting of that
enclosure to a control  device. A good enclosure
with an exhaust system can provide essentially 100
percent capture. For open ships and barges which
use buckets and  conveyors, a  partially enclosed
bucket will reduce windblown dust. When observa-
tion of the  bucket by the operator is required, a
transparent heavy plastic sheet can be used as a
cover. This  system  is only partially effective and
must usually be supplemented with other controls,
such as tighter fitting covers, wind breaks, or possi-
bly wet suppression.
                                   3.3.2.4  Area Fugitive Emission Control From
                                          Paved and Unpaved Roads
                                   Dust on the surface of paved roads is deposited by
                                   such processes as mud track-out on vehicle tires,
                                   atmospheric fallout, spillage or leakage from
                                   trucks,  pavement wear and decomposition, runoff
                                   or wind erosion from adjacent land  areas, deposi-
                                   tion of  biological debris, wear from tires and brake
                                   linings, and wear of anti-skid compounds. This ma-
                                   terial is reentrained by contact with tires and by the
                                   air turbulence created by passing vehicles.
                       42

-------
Table 3-8.    Control Technology Applications for Loading and Unloading Operations (16)

                                                       Control Procedures
Emission Points
            Loading Operations
          Unloading Operations
Railcar, Truck
Barge, Ship
-Drive-through enclosure with doors at both
 ends
-Exhaust of entire enclosure to dust removal
 equipment                       ,
-Movable hood over hatch opening
-Exhaust of car hopper to dust removal
 equipment
-Choke-feed or telescopic chute to confine and
 limit free-fall distance (gravity loading)
-Wet suppression (water, chemicals)
-Use of tarpaulins or covers over the holds
-Canopy and exhaust system over the loading
 boom, with attached tarps around the hatch
-Exhaust of ship hold to dust removal equipment
-Choke-feed or telescopic chute to confine and
 Mm it free-fall distance
-For tanker types, use of gravity filler spouts with
 concentric outer exhaust duct to control
 equipment
-Wet suppression (water, chemicals)
— Drive-through enclosure with doors at both
  ends
—Exhaust of enclosure to dust removal
  equipment
—Exhaust air from below grating of receiving
  hopper to removal equipment
—Choke-feed to receiving pit (hopper car and
  hopper truck)
—Unloading with screw conveyor (box car)
—Wet suppression (water, chemicals)
—Use of pneumatic unloading system

—Enclosure of top of clamshell bucket with
  transparent material and maintenance of
  closure seals and teeth on bottom of bucket
—Enclosure of shoreside receiving hopper
—Exhaust of enclosed shoreside receiving hopper
  to dust removal equipment
On unpaved roads, the road base itself serves as
the main source of dust. As with paved roads, the
dust becomes airborne by  contact with  vehicles'
tires and by air turbulence from passing vehicles.
Also, some of the fugitive dust from unpaved roads
is attributed to wind erosion. On both paved and
unpaved roads, traffic  movement causes the con-
tinuing mechanical breakdown of large particles on
the road surface, thus providing new material in
the  suspended  particulate  size range.  Available
procedures  for  reducing  emissions from plant
roads and their estimated efficiencies are present-
ed in Table 3-9.

Paved streets and  roads in a plant area  can  be
cleaned on  a frequent schedule  to reduce the
amount of particulate material on the surface that
is available for reentrainment. Flushers  and vacu-
um-type motorized street cleaners are both quite
effective in removing surface material and thereby
reducing emission  rates from vehicles  using the
cleaned streets. Because raw material accumulates
rapidly on the streets, the overall effectiveness of a
street-cleaning  measure is  a function  of the fre-
quency  of cleaning and the removal efficiency of
the equipment.
For plants with small amounts of paved roads, in-
dustrial vacuum sweepers or contracted sweeping
programs  (such as many  shopping  centers use)
would be more appropriate than the larger vacuum
street cleaning equipment  used on public streets.
Mechanical broom sweepers have been shown to
be ineffective from an  air pollution control stand-
point in  that they redistribute material into the ac-
tive  traffic lanes of the streets and  they remove
almost  none of the fine material (less than 43 |xm)
that is subject to reentrainment.
                             Many street sweepers depend  upon the material
                             being concentrated in the gutter in order to achieve
                             good collection efficiency, and therefore cannot be
                             used on streets without curbs and gutters. Howev-
                             er, the smaller industrial sweepers are usually de-
                             signed for use in warehouse and storage areas that
                             are not curbed. A factor which might limit the appli-
                             cability of street flushers in plants is that unpaved
                             areas adjacent to the streets would be wet by the
                             water spray and then become subject to mud track-
                             out onto the streets  by equipment and vehicles
                             driving through these areas.

                             Good housekeeping practices include the rapid re-
                             moval of spills on roadways and at conveyor trans-
                             fer points. Preventive measures include covering of
                             truck beds to prevent windage  losses,  cleaning of
                             truck tires and undercarriages to reduce mud track-
                             out onto paved roads, and minimizing  the pick-up
                             of mud by trucks.
                             Table 3-9.   Control Technology Applications for
                                        Plant Roads (16)
                              Emission Points
        Control Procedure
Efficiency
                              Paved Streets
                              Unpaved
                              Roads
                              Road
                              Shoulders
   Street cleaning               No estimate
   Housecleaning programs to
     reduce deposition of material
     on streets                 No estimate
   Vacuum street sweeping
     (daily) (2)                   25% (17)
   Speed reduction               Variable

   Paving                       85%
   Chemical stabilization            50%
   Watering                     50%
   Speed reduction               Variable
   Oiling and double chip surface     85%
   Stabilization                    80%
                                                                            43

-------
The paving of unpaved roadways is the most per-
manent of the various types of controls. However,
the degree  of effectiveness of this technique  is
highly dependent on prevention of  excessive sur-
face dust loading.

Watering of unpaved roads is effective only when
carried out on a regular basis. The schedule de-
pends on climate, type of surface material, vehicle
use, and type of vehicles.

Oiling unpaved roads is more effective than water-
ing and needs to be applied less often. However,
special precautions must always be taken so as not
to add to surface water runoff problems.
3.3.2.5 Area Fugitive Emission Control From
       Storage Piles
Most dust arises from stockpile areas as the materi-
al is dumped from the conveyor or chute onto the
pile, and as bulldozers move the pile. During per-
iods with high wind speeds (greater  than about 6
m/sec [13 mph]) or low moisture, wind erosion of a
nonweathered surface may also cause emissions.
Applicable control  techniques for open storage
piles are presented in Table 3-10.

Enclosing materials in  storage is generally the
most effective means of reducing emissions from
this source category because it allows  the emis-
sions to  be captured. However, storage bins or si-
los may be very expensive. Storage buildings must
be designed to withstand wind and snow loads and
to meet  requirements for interior working  condi-
tions. One alternative to enclosure of all material is

Table 3-10.   Control Technology Applications for
           Open Storage Piles (16)
Emission Points Control Procedure
Loading onto Enclosure
Piles Chemical wetting agents or
foam
Adjustable chutes
Movement of Enclosure
Pile Chemical wetting agents
Watering
Traveling booms to distribute
material
Efficiency
70-99%
80-90%
75%
95-99%
90%
50%
No estimate
Wind Erosion  Enclosure                   95-99%
            Wind screens                Very low
            Chemical wetting agents or
             foam                      90%
            Screening of material prior to
             storage, with fines sent
             directly to processing or to a
             storage silo               No estimate

Loadout      Water spraying                50%
            Gravity feed onto conveyor       80%
            Stacker/reclaimer             25-50%
to screen the material prior to storage, sending the
oversize material to open storage and the fines to
silos.

Wind screens, or partial enclosure of storage piles,
can reduce wind erosion losses but do not permit
capture of the remaining storage pile fugitive emis-
sions. Earthen berms, vegetation, or existing struc-
tures can serve as wind screens.

Telescoping chutes, flexible chute extensions, and
traveling booms are used to minimize the free fall
of material onto the  pile and resulting emissions.
Similarly, emissions due to loadout can be reduced
by reclaiming the material from the bottom of the
pile with a mechanical plow or hopper system. The
use of telescoping chutes and flexible chute exten-
sions for piles  with high material flow rates may
require closer control  of operations because of the
possibility  of jamming.  Traveling  or adjustable
booms can handle high flow rates, but have greater
operating  costs.

Wetting agents or foams that are sprayed onto the
material during processing or at transfer points re-
tain their effectiveness in subsequent storage oper-
ations. Wetting agents retain surface moisture for
extended  periods,  thereby  preventing  dusting.
Spraying of the material prior to storage may not
be possible in cases where product contamination
could result (e.g.,  Portland cement  clinker)  or
where the material is water soluble. However, such
materials are generally not placed in open storage
anyway. Steam has also been found to be an effec-
tive dust suppressant for some short-term storage
operations.

3.3.2.6 Area Fugitive Emission Control From Waste
      Disposal Sites
Fugitive dust can occur anywhere dusty waste ma-
terial is dumped for disposal. This includes over-
burden  piles, mining  spoils, tailings, fly  ash, bot-
tom ash, catch from air pollution  control
equipment, process overload discharges, building
demolition wastes, contaminated product, etc. Like
open storage, emissions come from dumping and
from wind  erosion across unprotected  surfaces.
Since waste piles are generally  not disturbed after
dumping, there are no emissions from an activity
comparable to loading out of the storage pile. How-
ever, there may be emissions from transporting the
waste material on-site (if it  is dry when  it is pro-
duced) or from a reclamation process such as land-
fill covering associated with the waste disposal op-
eration. If the surface of the waste material does
not include a compound that provides cementation
upon weathering, or if the surface is not compact-
ed, or if an area of very little rainfall, wind erosion
of fines can occur with winds greater than about 21
km per hour (13 mph). Table 3-11 presents control
techniques for waste disposal sites.
                      44

-------
Table 3-11.   Control Technology Applications for Waste
           Disposal Sites (16)
Emission Points
Control Procedure
Efficiency
Handling      Keep material wet              100%
            Cover or enclose hauling       No estimate
            Minimize free fall of material    No estimate

Dumping      Spray bar at dump area          50%
            Minimal free fall of material     No estimate
            Semi-enclose bin            No estimate
Wind Erosion





Grading
Cover with dirt or stable
material
Chemically stabilize
Revegetate
Rapidly reclaim newly
filled areas
Water

100%
80%
25% -100%

No estimate
50%
3.4  References
 1. Control  Technology for Toxic and Hazardous
    Air Pollutants. McFarland, A. R., ed. Illinois In-
    stitute for Environmental Quality. Chicago, Illi-
    nois. 1975.

 2. Committee on Industrial Ventilation.  Industrial
    Ventilation: A Manual of Recommended Prac-
    tice. 17th Edition. Lansing, Michigan. 1982.

 3. U.S. EPA. Evaluation of Control Techniques for
    Hazardous  Air Pollutants -  Appendices. EPA-
    600/7-86-009D (NTIS PB86-167/038/AS.)October
    1985.

 4. U.S. EPA. Flexible Vinyl Coating and Printing
    Operations — Background Information for Pro-
    posed Standards. EPA-450/3-81-016a. January
    1983.

 5. U.S. EPA. Publication Rotogravure Printing —
    Background Information  for Proposed Stan-
    dards. EPA-450/3-80-031a. October 1980.

 6. U.S. EPA. Measurement of Process Capture Ef-
    ficiency.  Draft Report of Laboratory Testing.
    EPA Contract No. 68-03-3038. January 1984.

 7. U.S. EPA. VOC Emissions from Petroleum Re-
    finery Wastewater Systems — Background In-
    formation for Proposed Standards. Research
    Triangle Park, North Carolina. July 1984.
 8.  U.S. EPA. Assessment of Atmospheric Emis-
    sions from Petroleum Refining — Appendix B:
    Detailed Results.  EPA-600/2-80-075c. April
    1980.

 9.  U.S. EPA. VOC Fugitive Emissions in Synthetic
    Organic Chemicals Manufacturing Industry —
    Background Information for Promulgated Stan-
    dards. EPA-450/3-80-033b. June 1982.

10.  U.S. EPA. Fugitive Emission Sources of Organ-
    ic Compounds —  Additional Information  on
    Emissions, Emission Reductions,  and  Costs.
    EPA-450/3-82-01 O.April 1982.

11.  Wilkins, G. E., J. H. E. Stelling, and S. A. Shar-
    eef. Monitoring and Maintenance Programs for
    Pumps and Valves in Petroleum and Chemical
    Processing Plants: Costs and Effects on Fugi-
    tive Emissions. Presented at 6th World Con-
    gress on Air Quality, IUAPPA, Paris. May 1983.

12.  Wilkins, G. E., and J. H. E. Stelling. Monitoring
    and Maintenance Programs for Control of Fugi-
    tive Emissions from Pumps and Valves in Pe-
    troleum and Chemical Processing  Plants. Pre-
    sented at 1984 Industrial Pollution  Control
    Symposium, ASME. New Orleans, Louisiana.
    February 1984.

13.  U.S. EPA. VOC Fugitive Emission Predictive
    Model — User's Guide. EPA-600/8-83-029.  Oc-
    tober 1983.

14.  U.S. EPA. Evaluation of the Efficiency of Indus-
    trial Flares. Background — Experimental  De-
    sign  — Facility.  EPA-600/2-83-070.  August
    1983.

15.  Environmental Engineers' Handbook, Volume
    II: Air Pollution. Liptak, B. G., ed. Chilton Book
    Company. Radnor,  Pennsylvania. 1974.

16.  U.S. EPA.  Technical Guidance for Control of
    Industrial Process  Fugitive  Paniculate  Emis-
    sions. EPA-450/3-77-010. March 1977.

17.  U.S. EPA. Identification, Assessment, and Con-
    trol of Fugitive Paniculate Emissions. Final re-
    port. EPA Contract No. 68-02-3922. May 1986. Con-
    tact Mr. Dale Harmon of EPA at (919) 541-2429.
                                                                       45

-------
                                          Chapter 4
                                  HAP Control Techniques
4.1 Background
This section describes and illustrates the proce-
dures used to calculate the basic design and oper-
ating variables of HAP control techniques in terms
of commonly employed design principles and val-
ues. For each technique,  the  handbook provides:
(a) a brief description of how the technique works,
(b) definitions of input data required, and (c) a step-
by-step calculation procedure showing where each
number used in the procedure comes from and
how it is to be used. The procedures described in
this  handbook  will result  in  conservatively  de-
signed control systems. In instances in which less
conservatively  designed  control  systems might
achieve the target control level, more detailed cal-
culation procedures requiring compound-specific
data would be  needed. This level of specificity is
beyond the scope of this handbook.

The data for the HAP emission stream to  be con-
trolled are  taken from  the HAP Emission  Stream
Data Form  given in Chapter 2. In case of a permit
evaluation, however, these data should be supplied
by the applicant. The reviewer may wish to confirm
the completeness of the applicant's data by refer-
ring to Chapters 2 and 3.

The step-by-step calculation procedures are illus-
trated for each control technique using data based
on Emission  Streams  1  through 7  described  in
Chapter 3. In permit reviews, the calculated values
are compared to the values in the permit applica-
tion to determine the adequacy of the applicant's
proposed design.  Appendices  C.2 through C.11
contain blank calculation sheets to use in applying
the calculations described for each  control tech-
nique.  If control systems costs are  required, see
Chapter 5.
4.2 Thermal Incineration
Thermal incineration (Figure 4-1) is a widely used
air pollution control technique whereby organic va-
pors are oxidized at high temperatures. The most
important variables to consider in thermal inciner-
ator design are the combustion temperature and
residence  time because these design variables de-
termine the  incinerator's destruction efficiency.
Further, at a given combustion temperature and
residence time, destruction efficiency is also affect-
ed by the degree of turbulence, or mixing of the
emission stream and hot combustion gases, in the
incinerator. In addition, halogenated organics are
more difficult to oxidize then unsubstituted organ-
ics;  hence, the  presence of halogenated com-
pounds in the emission  stream requires higher
temperature and longer residence times for com-
plete oxidation. Thermal incinerators can achieve a
wide range of destruction efficiencies. This discus-
sion focuses on efficiencies of 98 to 99 + percent.

The  incinerator flue gases are discharged at  high
temperatures and  contain valuable heat energy.
Therefore, there is  a strong economic incentive for
heat recovery. Typical recovery methods include
heat exchange between  the flue gases and the
emission stream and/or combustion air and use of
the available heat  for process heat requirements
(e.g., recycling flue gases to the process, producing
hot water  or steam, etc.).  In most thermal inciner-
ator applications, the available enthalpy in the flue
gases is used for preheating the emission stream.
This discussion will be based on a thermal inciner-
ation system where the emission stream  is  pre-
heated.

The  incineration of emission streams containing
organic vapors with halogen or sulfur components
may create additional control requirements. For ex-
ample, if sulfur and/or chlorine are present in the
emission stream, the resulting flue gas will contain
sulfur  dioxide (S02)  and/or  hydrogen  chloride
(HCI). Depending upon the concentrations of these
compounds in the flue gas and the applicable regu-
lations, scrubbing  may be required to reduce the
concentrations of these compounds. The selection
and  design of scrubbing systems are discussed in
Section 4.7.

In this subsection, the calculation procedure will be
illustrated using  Emission Stream 1 described in
Chapter 3. Appendix C.3 provides worksheets for
calculations.
                                              47

-------
Figure 4-1.  Schematic diagram of a thermal incinerator system.
                                                 Emission Source
                                                                    Scrubber*
         Combustion Air*

      Supplementary Fuel
                                 -Stack
                                   "Required for specific situations.
4.2.1 Data Required
The  data  necessary to  perform the calculations
consist of HAP emission stream characteristics pre-
viously  compiled on the HAP Emission Stream
Data Form  and the required HAP control as deter-
mined by the applicable regulations.
                 Example Case
  Maximum flow rate, Qe =  15,000 scfm
  Temperature, Te = 120°F
  Heat content, he = 0.4 Btu/scf
  Oxygen content, O2 = 20.6%
  Moisture content, Me = 2%
  Halogenated organics: Yes

  Based on the control requirements for the emis-
  sion stream:
  Required  destruction efficiency, DE = 99%
If dilution air is added to the emission stream upon
exit from process, the data that will be used in the
calculations are the  resulting characteristics after
dilution.

In the case of permit review for a thermal inciner-
ator, the data outlined below should be supplied by
the applicant; the calculations in  this section will
then be used to check the applicant's values.

Thermal  incinerator system  variables at standard
conditions (70°F, 1 atm):
Reported destruction efficiency, DEreported, %
Temperature of the emission stream entering the
incinerator, Te, °F (if no heat  recovery);
The, °F (if a heat exchanger is employed)
Combustion temperature, Tc °F
Residence time, tr, sec
Maximum emission stream flow rate, Qe, scfm
Excess air, ex, %
Fuel heating value (assume natural gas), hf, Btu/scf
Supplementary heat requirement, Hf, Btu/min
Combustion chamber volume, Vc, ft
Flue gas flow rate, Qfg, scfm
Heat exchanger surface area (if a  heat exchanger is
employed), A, ft2
4.2.2 Pretreatment of the Emission Stream:
     Dilution Air Requirements
In HAP emission  streams containing  oxygen/air
and flammable vapors, the concentration of flam-
mable vapors is generally limited to less than 25
percent of the lower explosive limit (LED to satisfy
safety requirements  imposed by insurance com-
panies. (Note: The LEL for a flammable vapor is
defined as the minimum concentration in air or
oxygen at and above which the vapor burns upon
contact with  an  ignition source and the flame
spreads through  the flammable gas mixture).  In
some cases, flammable vapor concentrations up to
40-50 percent of  the LEL are permitted if on-line
monitoring of VOC concentrations and automatic
process control and shutdown are provided.

In general, emission streams treated by thermal
incineration are dilute mixtures of VOC and air, and
typically do not require dilution.  For emission
streams with oxygen concentrations greater than
16 percent and heat contents greater than 13  Btu-
scf (corresponding to flammable vapor concentra-
tions  of approximately 25 percent of LEL), the cal-
culation procedure in this handbook assumes that
dilution air is required. (See Appendix B.2 for calcu-
lation  of dilution air requirements  and Appendix
C.2 for a calculation worksheet.)
                      48

-------
                 Example Case
  Since 02 = 20.6% and he =  0.4 Btu/scf, no dilu-
  tion air is required.
4.2.3  Thermal Incinerator System Design
      Variables
Table 4-1 presents suggested combustion tempera-
ture and residence time values for thermal inciner-
ators to achieve a given destruction efficiency. Two
sets of values are  shown in the table, one set for
nonhalogenated emission streams and another set
for halogenated emission streams. The combus-
tion temperature and residence time values listed
are conservative and assume adequate mixing of
gases in the incinerator. The criteria in this table are
not the only conditions for achieving the specified
destruction efficiencies. For a given destruction ef-
ficiency, it may be possible to incinerate HAP emis-
sion streams at lower temperatures with longer
residence times.

Based on the required destruction  efficiency (DE),
select appropriate valuesforTc and tr from Table 4-1.

Table 4-1.   Thermal Incinerator System Design
          Variables (1,2)
 Required  Nonhalogenated Stream   Halogenated Stream
Destruction Combustion  Residence  Combustion Residence
 Efficiency Temperature   Time   Temperature  Time
   DE(%)     TC(°F)     tr(sec)     Tc (°F)    tf (sec)
98
99
1,600
1,800
0.75
0.75
2,000
2,200
1.0
1.0
                 Example Case
 The required destruction efficiency is 99% and
 the  HAP emission  stream is  nonhalogenated,
 therefore:

 Tc   = 1,SOOT (Table 4-1}
 tr   = 0.75 sec (Table 4-1)
 In a permit evaluation, if the reported values for Tc
 and tr are sufficient  to achieve the required DE
 (compare the applicant's values with  the  values
 from  Table  4-1), proceed with the calculations. If
 the reported values for Tc and tr are not sufficient,
 the applicant's design is unacceptable. The review-
 er may then wish to  use the values for Tc and tr
 from Table 4-1. (Note: If DE is less than 98 percent,
 obtain information from literature and incinerator
 vendors to determine appropriate values for Tc and
4.2.4  Determination of Incinerator Operating
      Variables
4.2.4.1  Supplementary Heat Requirements
Supplementary fuel is added to the thermal inciner-
ator to attain the desired combustion temperature
(Tc). For  a given  combustion temperature, the
quantity of heat needed  to maintain the combus-
tion temperature in the thermal incinerator is pro-
vided by:  (a) the heat supplied from the combus-
tion of supplementary fuel, (b) the heat generated
from the combustion of hydrocarbons in the emis-
sion stream, (c) the sensible heat  contained in the
emission stream as it leaves the emission source,
and (d) the sensible heat gained  by the emission
stream through heat exchange with hot flue gases.
Typically,  items (b) and (c) are very small and can
often be neglected.

In general, emission  streams  treated by thermal
incineration are dilute mixtures of VOC and air, and
typically do not require additional combustion air.
For preliminary calculations, it can  be assumed that
no additional combustion air is required if the oxy-
gen concentration  of the emission stream exceeds
16 percent. Depending on the  heat content of the
emission stream and the  desired combustion tem-
perature, combustion air requirements may be zero
even when the oxygen concentration  is below 16
percent. Hence, this cut-off value will lead to a con-
servative  design.  Use the  following simplified
equation for dilute  streams to calculate supplemen-
tary heat requirements (based on natural gas):
                                                  Hf = 1.1 hf

                                                         f 0.002 Me) [CPair(Tc-Tr) - Cpair(The-Tr) - he]
                                                                hf-1.4Cpair(Tc-Tr)
                                                                                          (4.2-1)
where:

Hf   =
hf   =
Qe   =
Me   =
        supplementary heat requirement, Btu/min
        heating value of natural gas, Btu/scf
        maximum emission stream flow rate, scfm
        moisture content of the emission stream, %
Cpair =  average specific heat of air over a given
        temperature interval (Tr to T), Btu/scf-°F
Tc   =  combustion temperature, °F
Tr   =  reference temperature, = 70°F
The  =  emission stream temperature after heat
        recovery, °F
he   =  heat content of the emission stream,
        Btu/scf

Calculate The using the following expression  if the
value for The is not specified:

The =  (HR/100)TC + [1 - (HR/100)] Te

where:
HR =  heat recovery in the exchanger, %

Assume a value of 50 percent if no other informa-
tion is available.
                                                                       49

-------
The  factor 1.1  in Equation 4.2-1  accounts for an
estimated heat loss of 10 percent in the incinerator.
Supplementary heat  requirements are  based on
maximum emission  stream flow rate, and hence
will lead to a conservative design. In this handbook,
it is assumed that the minimum supplementary
heat requirement is 5 Btu/min per scfm of emission
stream.

A graph of Equation 4.2-1 is  shown in Figure 4-2,
where the ratio Hf/Qe is plotted against the emis-
sion stream heat content (he) for four different com-
bustion temperatures  (Tc). Instead of evaluating
Equation 4.2-1, Figure 4-2 can be used directly to
determine supplementary heat requirements. This
graph is based on the following assumptions: (1)
temperature of the emission  stream (Te) is 100°F,
(2) moisture content of the emission stream (Me) is
2 percent, (3) preheat temperature of the emission
stream (The) is based on 50 percent heat recovery in
the heat exchanger, and (4) hf = 882 Btu/scf (based
on the lower heating value for natural  gas).
Figure 4-2.  Supplementary heat requirement vs. emission
          stream heat content (dilute stream/no
          combustion air).
                                                  Figure 4-3.
                                                       100
^_   90

II
ffi~»   80
          Supplementary heat requirement vs. emission
          stream heat content (no oxygen in emission
          stream/maximum combustion air).
                                                  s£
                                                  1 E
                                                  I"
                                                  E o
      70


      60


      50


      40


      30


      20
                                                               • Tc = 2,200°F
                                                               O Tc = 2,000°F
                                                               • Tc = 1,800°F
                                                               D Tc = 1,600°F
                                                                  1234

                                                              Emission Stream Heat Content, he (Btu/scf)
     50
  I
  O  40
I
  C  30
  E
  • ro
   o
 « •«
 Q.  10
 §E
05 W
     20
     10
           • Tc = 2,200°F
           O Tc = 2,000°F
           • Tc = 1,800°F
           D Tc = 1,600°F
               1234
           Emission Stream Heat Content, he (Btu/scf)
For emission streams that do not contain sufficient
quantities of oxygen to satisfy the combustion air
requirements (e.g., process emissions), refer to Fig-
ure 4-3 which shows a plot of Hf/Qe versus he to
obtain a conservative estimate for Hf. Figure 4-3 is
based on the same assumptions as those stated for
Figure 4-2. In addition, the oxygen  content of the
emission stream (02) is assumed  as zero; this cor-
responds  to maximum  combustion  air require-
ments for the thermal incinerator  system.  If the
oxygen content of a  particular emission stream
falls between zero and  16 percent, use Figure 4-3 to
obtain a conservative estimate of  Hf/Qe.
                  Example Case
  Using Equation 4-1:
  Since the emission stream is very dilute and has
  an oxygen content greater than 16%, Equation
  4.2-1 is applicable. The values to be inserted in
  the equation are:

       =  15,000 scfm
       =  2%
       =  0.4 Btu/scf
       =  1,800°F (Table 4-1)
       =  70°F
       =  960°F (based on heat recovery of 50%)
       =  0.0196 Btu/scf-°F for the interval
          70° - 1,800°F (reference 3)
  Cpair =  0.0187 Btu/scf-°F for the interval
          70° - 960°F (reference 3)
  hf   =  882 Btu/scf
                                                    Me
                                                    he
                                                    To
                                                    Tr
                                                    The
                                                    Hf   = 1.1x882
                                                                     "[15,000(33.91 -16.64-0
                                                                             882 - 47.5
                                                                                           •4H1
                                                    Hf    = 295,000 Btu/min

                                                    (Note:  Hf is greater than the minimum supple-
                                                    mentary heat requirement assumed in this hand-
                                                    book.)

                                                    Using Figure 4-2:
                                                    For he  =  0.4 Btu/scf and Tc  = 1,800°F, Hf/Qe as
                                                    indicated in the figure is about 20 Btu/min/scfm.
                                                    Multiplying 20 by Qe, (20 x 15,000),  yields an
                                                    approximate value of 300,000 Btu/min for Hf.
                      50

-------
4.2.4.2 Flue Gas Flow Rate
Flue gas is generated as a result of the combustion
of supplementary fuel and the emission  stream.
Use the following equation to calculate flue gas
flow rate:

Qfg = Qe + Qf +  Qc                       (4.2-2)

where:
Qfg  = flue gas flow rate, scfm
Qf   = natural gas flow rate, scfm
Qc   = combustion air requirement, scfm

Calculate Qf from the following equation:

Of = Hf/hf                                (4.2-3)

As  indicated earlier, Qc will typically be zero for
dilute emission streams with oxygen contents (02)
greater than 16 percent. If 02 is less than 16 per-
cent, then:

Qc = (0.01  He + 9.4 Qf) (1 + 0.01 ex) - 0.0476 02 Qe

where:
He  = heat generated due to the combustion  of
      hydrocarbons in the emission stream
He = Qehe
                 Example Case
  Using Equation 4.2-3:
  Hf  = 295,000 Btu/min
  h,  = 882 Btu/scf
  Qf  = 295,000/882 = 330 scfm

  Using Equation 4.2-2:
  Qe   = 15,000 scfm
  Qc   = 0 (since 02 is greater than 16%)
  Qfg  = 15,000 + 330 + 0
  Qfg  = 15,330 scfm
4.2.5 Combustion Chamber Volume
The flue gas flow rate (Qfg) determined by Equation
4.2-2  is expressed at standard conditions. In order
to calculate the combustion chamber size, Qfg must
be expressed at actual conditions, i.e., temperature
effects  must be considered (assume pressure ef-
fects  are negligible). Use the following equation to
convert Qfg from "scfm" to "acfm":

Qfg,a = Qfg KTC +  460)7530]                 (4.2-4)
where:
Qfg.a = flue gas flow rate at actual conditions, acfm

The volume of the combustion chamber (Vc) is de-
termined from the residence time (tr) and flue gas
flow rate  at actual  conditions (Qfg,a) according to
the following equation:
         The factor 1.05 is used in Equation 4.2-5 to increase
         the chamber volume by 5 percent. This technique is
         an accepted industry practice and allows for fluctu-
         ations in the operating conditions (e.g., flowrate,
         temperature,  etc.).  The  smallest  commercially
         available incinerator has  a combustion chamber
         volume of about 36 ft3 (1 m3). If the calculated Vc is
         less than 36 ft3, define Vc as 36 ft3.
                          Example Case
          Using Equation 4.2-4:
          Tc   =  1,800°F
          Qfg  =  15,330 scfm
          Qfg,a =  15,330 [(1,800 + 460)/530]
          Qfg'a =  65,370 acfm

          Using Equation 4.2-5:
          tr    =  0.75 sec (from Table 4-1)
          Vc   =  [(65,370/60)0.75] 1.05
          Vr   =  860 ft3
         4.2.6 Heat Exchanger Size
         The size of the  heat exchanger required for pre-
         heating the emission stream to The before it enters
         the thermal incinerator is based on the heat ex-
         changer design. Use the following expression to
         calculate the required size, i.e., surface area, of the
         heat exchanger:
         A =
              [60Qe(1 + 0.002 Me) Cpair (The - Te)]
                           UAT,
                                         (4.2-6)
                               LM
                  heat exchanger surface area, ft2
                  emission stream temperature, °F
                  overall heat transfer coefficient, Btu/hr-ft2-°F
                  logarithmic mean temperature difference, °F
Vc= [(Q,g,a/60)tr]x1.05
(4.2-5)
where:
A
Te    =
U
ATLM  =
and:
ATUM  =  [(Tc - The) - (Thc - Te)] In [(Tc - The)/(Thc - Te)]

For dilute emission  streams that do not require
additional combustion air, ATLM can be approxi-
mated by:

ATLM  = Tc - The
For a  recuperative heat exchanger where the heat
transfer takes place between two gas streams, the
overall heat transfer coefficient (U) ranges from 2
to 8 Btu/hr-ft°F,  generally depending on the heat
exchanger configuration and properties of the gas
streams.

Equation 4.2-6 has been evaluated for dilute emis-
sion streams that require no additional combustion
air, as shown in Figure 4-4. In the figure, heat ex-
changer  surface area  (A)  is plotted against the
emission stream flow rate (Qe). The  assumptions
inherent  in this figure are the same  as  those de-
scribed for Figure 4-2. The overall  heat transfer
coefficient is assumed as 4 Btu/hr-ft2-°F.
                                                                       57

-------
                  Example Case
  Using Equation 4.2-6:
  Qe   =  15,000 scfm
  M
=  2%
 he
         960°F (based on heat recovery of 50%)
         120°F (input data)
         0.0187 Btu/scf-°F for the interval 120° -
         960°F (reference 3)
         4 Btu/hr-ft2-°F
         Tc-The
         1,800°F
  ATLM= 1,800°-960° = 840°F
  Substituting in Equation 4.2-6:
          [60 x 15,000 x 1.004 x 0.0187 (960 - 120)]
  A    = 	
U
ATLM =
Tr   =
  A    = 4,200ft2
                        (4 x 840)
  Using Figure 4-4:
  For Qe = 15,000, the value for A from the figure
  is about 4,000 ft2.
For emission  streams that  are not dilute and re-
quire additional combustion air, use Figure 4-5 to
obtain an estimate of the heat exchanger surface
area. In Figure 4-5, the ratio A/Qe is plotted against
the emission stream content (he) for four different
combustion temperature (Tc). The  assumptions in-
herent in this figure are the same as those  stated
for Figure 4-4; in  addition,  maximum combustion
air requirements are assumed (i.e., 02 = 0). If the
conditions represented in Figures 4-4 and 4-5 are
not directly applicable for  a particular emission
stream,  use  Figure  4-4 to  obtain  a  conservative
estimate.
                                                  If the calculated values and the reported values are
                                                  not different, then the design and operation of the
                                                  proposed thermal incinerator system may be con-
                                                  sidered  appropriate based on  the assumptions
                                                  used in this handbook.
                                             Figure 4-4.   Heat exchanger size vs. emission stream flow rate
                                                       (dilute stream/no combustion air).
                                                    15,000
                                                    12,500
                                                  N""
                                                  C
                                                  <

                                                  
-------
Table 4-2.   Comparison of Calculated Values and Values
          Supplied by the Permit Applicant for Thermal
          Incineration

                           Calculated
                              Value      Reported
                         (Example Case)a   Value
Supplementary heat
requirement, Hf
Supplementary fuel flow
rate, Q,
Flue gas flow rate, Qfg
Combustion chamber size, Vc
Heat exchanger surface area, A
295,000
330 scfm
1 5,330 scfm
860ft3
4,200 ft2
aBased on Emission Stream 1.
4.3 Catalytic Incineration
Catalytic incineration (Figure 4-6) is an air pollution
control  technique whereby VOC's in an emission
stream  are oxidized with the help of a  catalyst. A
catalyst is a substance that accelerates the rate of a
reaction at a given temperature without being ap-
preciably  changed during  the  reaction. Catalysts
typically used for VOC incineration include  plati-
num and palladium;  other formulations  are also
used, including metal oxides for emission streams
containing chlorinated  compounds.  The catalyst
bed  (or matrix) in the incinerator is generally a
metal mesh-mat, ceramic honeycomb, or other ce-
ramic matrix structure designed to maximize cata-
lyst surface area. The catalysts may also be in the
form of spheres or pellets.  Before passing through
the catalyst bed, the emission stream is  preheated,
if necessary, in a natural gas-fired preheater.

The performance of a catalytic incinerator  is affect-
ed by several factors including: (a) operating tem-
perature, (b) space velocity (reciprocal of residence
time), (c) VOC composition and concentration, (d)
catalyst properties, and (e) presence of poisons/in-
hibitors in the emission stream. In catalytic inciner-
ator design, the important variables are the operat-
ing temperature at the catalyst bed inlet and the
space velocity. The operating temperature  for a
particular destruction  efficiency  is dependent on
the concentration and composition of the VOC  in
the emission stream and the type of catalyst  used.

Space velocity  is defined  as the volumetric flow
rate of the combined gas stream (i.e., emission
stream + supplemental fuel  + combustion air) en-
tering the catalyst bed divided by the volume  of the
catalyst bed. As such, space velocity also depends
on the type of catalyst used. At a given space veloc-
ity, increasing the operating temperature at the in-
let of the catalyst bed increases the destruction
efficiency. At a given operating temperature, as
space velocity is decreased (i.e., as residence time
in the catalyst bed increases), destruction efficiency
increases. Catalytic incinerators can achieve overall
VOC destruction efficiencies up to about 98 percent
and HAP destruction efficiencies up to about 95
percent with space velocities in the range 30,000 to
100,000  hr1.(4,5,6) However, the greater catalyst
volumes and/or higher temperatures required for
higher  destruction efficiencies  (i.e., 99 percent)
may  make catalytic incineration  uneconomical.
This discussion will be based on HAP destruction
efficiencies of 90 and 95 percent.

The performance of catalytic incinerators is sensi-
tive to pollutant characteristics and process condi-
tions. In the following discussion,  it is assumed that
the emission stream is free from poisons/inhibitors
such as phosphorus, lead, bismuth, arsenic, anti-
mony, mercury, iron oxide, tin,  zinc, sulfur, and
halogens. (Note: Some catalysts can handle emis-
sion streams containing halogenated compounds.)
Figure 4-6.  Schematic diagram of a catalytic incinerator system.
                                                                 Emission Source
                                                                    O
      Combustion Air*
   Supplementary Fuel





Catalytic Incinerator
Preheater
I

»f V
**N
\A
M Vv
v^^
\-Catalyst Bed
1


*• otacK
Heat Exchanger
(Optional)
                                                                                   -Dilution Air*
                                     "Required for specific situations.
                                                                        53

-------
It is also assumed that the fluctuations in process
conditions (e.g., changes in VOC content) are kept
to a minimum.

The energy in the flue gases leaving the catalyst
bed may be recovered in several ways including:
(a) use of a recuperative heat exchanger to preheat
the emission  stream and/or combustion air, or (b)
by use of the available energy for process heat
requirements (e.g., recycling flue gases to the pro-
cess, producing hot water or steam, etc.). In recu-
perative  heat exchange, only a limited preheat is
possible due to the temperature rise  across the
catalyst bed as a result of the combustion of VOC in
the emission stream. High preheat temperatures
accompanied by an  increase in temperature due to
combustion result in high operating temperatures
at the catalyst bed, causing the catalyst bed to over-
heat and eventually  lose its activity.

The following discussion will be based on fixed bed
catalytic incinerator  systems with no heat recovery
and with recuperative heat exchange (i.e., preheat-
ing the emission stream). The calculation proce-
dure will be  illustrated  using  Emission Stream 2
described in  Chapter 3. Appendix C.4 provides
worksheets for calculations.
4.3.1 Data Required
The data necessary to perform  the  calculations
consist of HAP emission stream characteristics pre-
viously compiled  on the HAP Emission  Stream
Data Form and the required HAP  control as deter-
mined by the applicable regulations.
                 Example Case
 Maximum flow rate, Qe = 20,000 scfm
 Temperature, Te = 120°F
 Heat content, he = 2.1 Btu/scf
 Oxygen content, 02 = 20.6%
 Moisture content, Me = 2%

 Based on the control requirements for the emis-
 sion stream:
 Required destruction efficiency, DE = 95%
If dilution air is added to the emission stream upon
exit from the process, the data that will be used in
the calculations are the resulting characteristics
after dilution.

In the case of a permit review for a catalytic inciner-
ator, the following data should be supplied by the
applicant. The calculations in this section will then
be used to check the applicant's values.
Catalytic incinerator system variables at standard
conditions (70°F, 1 atm):

Reported destruction efficiency, DEreportedi %
Temperature of the emission stream entering the
incinerator, Te, °F (if no heat recovery); The, °F (if
emission stream is preheated)
Temperature of flue gas leaving the catalyst bed,
T   °F
 I co»  r
Temperature of combined gas stream  (emission
stream  + supplementary fuel  combustion  prod-
ucts) entering the catalyst bed,3 TCI, °F
Space velocity, SV, hr"1
Supplementary heat requirement, Hf, Btu/min
Flow  rate of combined gas  stream entering  the
catalyst bed,3 Qcom, scfm
Combustion air flow rate, Qc, scfm
Excess air, ex, %
Catalyst bed requirement, Vbed, ft3
Fuel heating value, hf, Btu/scf
Heat exchanger surface area (if a heat exchanger is
employed), A, ft2


4.3.2 Pretreatment of the Emission Stream:
     Dilution Air Requirements
In  general, catalytic incineration is applied to dilute
emission  streams. If emission  streams  with  high
VOC concentrations (i.e., heat content  above  10
Btu/scf for air + VOC  mixtures and  above  15
Btu/scf for inert  + VOC mixtures)  are treated  by
catalytic incineration, they may generate  enough
heat upon combustion to deactivate the catalyst.
Therefore, dilution of the emission stream  with air
is  necessary to reduce the concentration of the
VOC's.

Typically, the concentration of flammable vapors in
HAP emission streams containing air is limited to
less than 25 percent of the LEL (corresponding to a
heat content of 13 Btu/scf) for safety requirements.
In  order to meet  the safety  requirements and to
prevent damage to the catalyst bed, it is assumed
in  this handbook that catalytic incineration  is  appli-
cable if the heat content of the emission stream (air
 +  VOC) is less than or equal to 10  Btu/scf. For
emission  streams that are mixtures of inert gases
and VOC (i.e., containing no oxygen), it is assumed
that  catalytic incineration is applicable if the heat
content of the emission stream is less than  or equal
to 15 Btu/scf. Otherwise, dilution  air will be re-
quired to reduce  the heat content to levels below
these cut-off values (i.e., 10 and 15 Btu/scf). For
emission  streams that cannot be characterized as
air + VOC or inert + VOC mixtures, apply the 10
"If no supplementary fuel is used, the value for this variable will be the
 same as that for the emission stream.
                      54

-------
Btu/scf cut-off value for  determining dilution air
requirements. See Appendix B.2 for calculating di-
lution air requirements.
Based on the required destruction efficiency (DE),
specify the appropriate  ranges for Tci, Tc
select the value for SV from Table 4-3.
and
                  Example Case
  Since the  heat content of the emission stream
  (he) is 2.1 Btu/scf, no dilution is necessary.
4.3.3 Catalytic Incinerator System Design Variables
Table 4-3 presents suggested values and limits for
the design variables of a fixed bed catalytic inciner-
ator system to achieve a given destruction efficien-
cy. For specific applications, other temperatures
and space velocities may be appropriate depend-
ing on the type of catalyst employed and the emis-
sion stream characteristics (i.e., composition  and
concentration).  For  example, the temperature of
the flue gas leaving  the catalyst bed may be lower
than 1,000°F for emission streams containing easily
oxidized compounds and still achieve the desired
destruction efficiency. (See reference 7 or Appen-
dix B.6, reference 8, for data on temperatures typi-
cally required for specific destruction efficiency lev-
els for several compounds.)
Table 4-3.    Catalytic Incinerator System Design
           Variables (1,5,6)
Required
Destruction
Efficiency
DE(%)
90
95
Temperature
at the Catalyst
Bed Inlet3
Tc, (°F)
600
600
Temperature
at the Catalyst
Bed Outletb
Tco (°F)
1,000-1,200
1,000-1,200
Space
Velocity
SV(hr1)
40,000C
30,000d
a Minimum temperature of combined gas stream  (emission
 stream + supplementary fuel combustion products) entering
 the catalyst bed is designated as 600°Fto ensure an adequate
 initial reaction rate.
bMinimum temperature of flue gas leaving the catalyst bed is
 designated  as 1,000°F to ensure an adequate overall reaction
 rate to achieve the required  destruction efficiency. Note that
 this is a conservative value;  it is in general a function of the
 HAP concentration (or heat content) and a temperature lower
 than 1,000°F may be sufficient to achieve the required destruc-
 tion level. Maximum temperature of flue gas leaving the cata-
 lyst bed is limited to 1,200°F to prevent catalyst deactivation by
 overheating.
•"Corresponds to 1.5 ft3 of catalyst per 1,000 scfm of combined
 gas stream.
dCorresponds to 2.0 ft3 of catalyst per 1,000 scfm of combined
 gas stream.
Note  that  the  destruction  efficiency  for a  given
compound may vary  depending  on whether the
compound is the only VOC in the emission stream,
or it is part of a mixture of VOC's.(4) The destruc-
tion efficiency  for a given compound in different
VOC mixtures may also vary with mixture composi-
tion. (See reference 4 or Appendix B.6, reference 8,
for compound-specific destruction efficiency data
for two different VOC mixtures.)
                  Example Case
  The  required destruction  efficiency is 95%;
  therefore:
  Tci (minimum) = 600°F
  Tco (minimum) = 1,000°F
  Tco (maximum) = 1,200°F
  SV = 30,000 hr1
In a permit evaluation,  determine if the reported
values  for  Tci, Tco,  and  SV  are appropriate to
achieve the required destruction efficiency by com-
paring the  applicant's values with the values in
Table 4-3. The reported  value for Tci should equal
or exceed SOOT in order to obtain an adequate
initial reaction rate. To  ensure that an adequate
overall  reaction rate  can be achieved to give the
desired destruction efficiency without  damaging
the catalyst, check whether Tco falls in the interval
1,000° - 1,200°F. Note that 1,000°F is a conservative
value. Then check if the reported value for SV is
equal to or less than the value in Table 4-3. If the
reported values are appropriate, proceed with the
calculations. In some  cases it may be possible to
achieve the desired destruction efficiency at a low-
er temperature level. Otherwise, the applicant's de-
sign is considered unacceptable. In such  a case, the
reviewer may then wish to use the values  in Table
4-3.

4.3.4  Determination of Incinerator System
      Variables

4.3.4.1 Supplementary Heat Requirements
Supplementary fuel is added to the catalytic incin-
erator system to provide the heat necessary to
bring the emission stream up to the required cata-
lytic oxidation temperature (TCI) for the desired lev-
el of destruction efficiency.  For  a given  TCI, the
quantity of supplementary heat needed is provided
by: (a) the  heat supplied from the combustion of
supplementary fuel, (b) the sensible heat contained
in the emission  stream as  it  enters the  catalytic
incinerator system, and (c) the sensible heat gained
by the  emission  stream through heat  exchange
with hot flue gases. If recuperative heat exchange
is not practiced at a  facility, then item  (c) will be
zero.

As mentioned earlier,  emission streams  treated by
catalytic incineration  are dilute  mixtures  of VOC
and air, and  typically do  not require  additional
combustion air. As a conservative cut-off  value, it
can be assumed  that no additional combustion air
is required  if the emission stream oxygen content
(O2) is greater than or  equal to 16  percent.
                                                                          55

-------
Before calculating the supplementary heat require-
ments, the temperature of the flue gas leaving the
catalyst bed (Tco) should  be estimated to ensure
that an adequate overall reaction  rate can  be
achieved to give the  desired destruction efficiency
without damaging the catalyst. In other words,
check whether Tco falls  in the interval  1,000°  -
1,200°F. Use the following expression to calculate
Tco, taking into consideration the temperature rise
across the catalyst bed due to heat generation from
combustion of VOC in the emission stream:
    = Tci + 50 he
(4.3-1)
where:
he = heat content of the emission stream, Btu/scf

In this expression, it is assumed that the heat con-
tent of the emission stream and the combined gas
stream is the same. Inserting Tci =  600°F, if Tco is
in the range 1,000° -  1,200°F, then Tci  = 600°F is
satisfactory. If  Tco is less than 1,000°F, use the
following equation to determine an  appropriate
value for Tci (above 600°F) and use this value in the
following calculations:
Tci =  1,000-50he
(4.3-2)
(Note: Emission streams with high heat contents
will  be diluted  based on  the  requirements dis-
cussed in Section 4.3.2. Therefore, values for Tco
exceeding 1,200°F should not occur.)

To  calculate supplementary  heat requirements
(based on natural gas as the fuel), use the following
simplified equation  for dilute  emission streams
that require no additional combustion air:

Hf =  1.1 hfxQe(1  +  0.002 Me)
      [Cpair (Tci-Tr) - Cpair (The - Tr)
                  ir (Tci - Tr)
                                          (4.3)
where:


h,f   =
0-e   =
MP   =
        supplementary heat requirement, Btu/min
        heating value of natural gas, Btu/scf
        maximum emission stream flow rate, scfm
        moisture content of the emission stream,
        percent
Cpair =  average specific heat of air over a given
        temperature interval, (Tr -T) Btu/scf-°F
Tci   =  temperature of combined gas stream
        entering the catalyst bed, °F
Tr   =  reference temperature, = 70°F
The  =  emission stream temperature after heat
        recovery, °F

Note that for the case of no heat recovery, The =
Te.  The factor 1.1 accounts for an estimated  heat
loss of  10 percent in the incinerator. Supplemen-
tary heat requirements are based  on  maximum
emission flow rate, and hence will lead to a conser-
vative design. In contrast to thermal incineration,
                                                  there is no minimum supplementary heat require-
                                                  ment specified for catalytic  incineration since no
                                                  fuel is needed for flame stabilization. Depending on
                                                  the VOC concentration, emission stream tempera-
                                                  ture, and  level  of heat recovery,  supplementary
                                                  heat requirements may be zero when heat recovery
                                                  is practiced.

                                                  A  plot of  Equation 4.3-3 is  shown in  Figure 4-7
                                                  where the ratio of Hf/Qe is plotted against the emis-
                                                  sion  stream heat content for two levels of heat
                                                  recovery (zero and 50 percent). As an alternative to
                                                  Equation 4.3-3, use Figure 4.7 directly to determine
                                                  Hf. The figure is based on the following assump-
                                                  tions: (1) moisture content of the emission stream
                                                  (Me) is 2 percent,  (2) emission stream temperature
                                                  (Te) is  100°F, (3) preheat temperature of the emis-
                                                  sion stream (The) is based on 50 percent heat re-
                                                  covery in  the heat exchanger, and (4) net heating
                                                  value of supplementary fuel (natural  gas)  is 882
                                                  Btu/scf.
         Figure 4-7.
              28
              26 -
          O   24
                                                            Supplementary heat requirement vs. emission
                                                            stream heat content (dilute stream/no
                                                            combustion air).
          CO
          tr

          I
          U_
          E
          co
          CD
          •*-*
          C/3

          O
          '
                                                      22
                                                      20
                                                       18
                                                       16
c   14
E

m   12
          x"
          

          I    4
          a
                           > No heat recovery
                           I 50% heat recovery
                        12345
                     Emission Stream Heat Content, he (Btu/scf)
                      56

-------
                  Example Case
  Using Equations 4.3-1, -2, and -3:
  Since the emission stream  is dilute (he =  2.1
  Btu/scf) and has an oxygen concentration great-
  er than 16% (O2 = 20.6%), these equations  are
  applicable.
  a.
Determine  if Tco
1,200°F:
                   falls in the range 1,000° -
     he
     Tco
        = 600°F
        = 2.1 Btu/scf (input data)
        = 600 +  (50x2.1) = 705°F
Since Tco is less than 1,000°F, use Equation
4.3-2 to calculate an appropriate value for Tci:

   Tci   = 1,000-(50x2.1) = 895°F

b. Determine  Hf (assume  recuperative heat
   recovery will be employed):
     Qe
     Me
     Tr
     The
     Cpair
     Hf   =
               (882 - 21.60)
     Hf   = 150,500 Btu/min

  Using Figure 4-7:
  For he = 2.1  Btu/scf and using the curve for
  50% heat recovery, Hf/Qe from the figure  is
  about 7.5 Btu/min/scfm. Multiplying 7.5 by Qe,
  (7.5 x 20,000), yields an approximate value of
  150,000 Btu/min for Hf.
          20,000 scfm
          2%
          70°F
           550°F (based on heat recovery of 50%)
           0.0187 Btu/scf-°F for the interval
           70° - 895°F (reference 3)
   Cpair =  0.0183 Btu/scf-°F for the interval
           70° - 550°F (reference 3)
           1.1 x 882 x 20,000 x 1.004
             (15.43 - 8.78)
o
of
ro
CC

^
u_
e
CO

                                                  a
                                                  a

                                                  CO
                                                 Figure 4-8.  Supplementary heat requirement vs. emission
                                                           stream heat content (no oxygen in emission
                                                           stream/maximum combustion air).
42

40

38

36

34

32

30

28

26

24

22
    18

    16

    14

    12

    10

     8

     6

     4

     2
                                                                                          12
        1 No heat recovery
         Heat recovery
         (The  =  550°F)
                                                                1234

                                                            Emission Stream Heat Content, he (Btu/scf)
For emission streams that do not contain sufficient
quantities of oxygen to satisfy combustion air re-
quirements  (e.g., process emissions), refer to Fig-
ure 4-8 which shows a plot of Hf/Qe versus he for
two levels of heat recovery (i.e., no heat recovery
and where The is 550°F).  In this figure, the oxygen
content of the emission stream (O2) is assumed as
zero; this corresponds to  maximum combustion air
requirements. The emission stream moisture con-
tent (Me), emission  stream temperature (Te)  and
the fuel heating  value (hf) are as specified for Fig-
ure 4-7. The preheat temperature of the emission
stream (The) is 550°F for  the heat recovery case. If
02 for a particular emission stream  is between 0
and 16 percent, use Figure 4-8.
                                                 4.3.4.2 Flow Rate of Combined Gas Stream
                                                        Entering the Catalyst Bed

                                                 In  order  to  calculate the quantity  of catalyst re-
                                                 quired, the flow rate of the combined gas stream
                                                 (emission stream + supplementary fuel combus-
                                                 tion products) at the inlet to the catalyst bed has to
                                                 be determined. Use the following equation:
                                                              Qf + Qc
                                                                                       (4.3-4)
                                                 where:
                                                 QCom = flow rate of the combined gas stream, scfm
                                                 Qf   = natural gas flow rate, scfm
                                                 Qc   = combustion air requirement, scfm
                                                                         57

-------
Calculate Qf from the following expression:

Qf   = Hf/hf                             (4.3-5)

As indicated  earlier, Qc will typically  be zero for
dilute emission streams with oxygen contents (02)
greater than 16 percent. If 02 is less than 16 per-
cent, then:

Qc = (0.01 He + 9.4 Qf) (1 + 0.01 ex) - 0.476 02Qe

where:
He = heat generated due to the combustion of
       hydrocarbons  in the emission stream
He = Qehe
                 Example Case
 Using Equation 4.3-4:
 Qe   = 20,000 scfm
 Qc   =0 (since 02 is greater than 16%)
 Qcom = 20,000 + 170 + 0
 Qcom = 20,170 scfm
 Using Equation 4.3-5:
 Hf   = 150,500 Btu/min
 hf   = 882 Btu/scf
 Qf   =170 scfm
4.3.4.3 Flow Rate of Flue Gas Leaving the Catalyst
       Bed
In order to determine costs for incinerators and to
size a heat exchanger for preheating the emission
stream, the flow rate of flue gas leaving the catalyst
bed must be determined.

Assume  that the flow rate of the combined gas
stream entering the catalyst bed is approximately
equal to the flow rate of the flue gas leaving  the
catalyst bed at standard conditions. The volume
change across the catalyst bed due to the combus-
tion of the VOC in the mixed gas stream is small,
especially when dilute emission streams are treat-
ed. Therefore,

Qfg = Qcom
where:
Qfg = flow rate of the flue gas leaving the catalyst
      bed, scfm

While figuring costs, assume  that catalytic inciner-
ators are designed for a minimum Qfg of 500 scfm.
Therefore, if Qfg is less than  500 scfm, define  Qfg
as 500 scfm.

In order to determine operating costs, the flue gas
flow rate (Qfg) has to be expressed at actual condi-
tions. Use the following equation to convert  Qfg
from "scfm" to "acfm":
Qfg,a =
                460)/530]


                      58
(4.3-6)
                                                 where:
                                                 Qfg;a is the flue gas flow rate at actual conditions
                                                 (acfm)
                                                                  Example Case
                                                   Using Equation 4.3-6:
                                                   Qfg   = Qcom = 20,170 SCfm
                                                   Tco   - 1,000°F
                                                   Qfg,a  = [20,170(1,000 + 460)75301
                                                   Qfg,a  = 55,600 acfm
                                                 4.3.5 Catalyst Bed Requirement
                                                 The total volume  of catalyst required for a given
                                                 destruction efficiency is determined from the de-
                                                 sign space velocity as follows:
                                                 Vbed = 60 Qcom/SV

                                                 where:
                                                 vbed = volume of catalyst bed required, ft3
                                                  (4.3-7)
                                                                  Example Case
                                                   Using Equation 4.3-7:
                                                   Qcom = 20,170 scfm
                                                   SV   = 30,000 hr1 (Table 4-3)
                                                   Vbed  = 60x20,170/30,000
                                                   Vbed  = 40 ft3
                                                  4.3.6 Heat Exchanger Size (for Systems with
                                                       Recuperative Heat Exchange Only)
                                                  To determine the size  of the heat exchanger re-
                                                  quired for preheating the emission stream to The
                                                  use the following expression:

                                                  A = [60 Qe(1 + 0.002 Me) Cpair (The - Te)]/UATLM
                                                                                          (4.3-8)
                                                  where:
                                                  A   = heat exchanger surface area, ft2
                                                  U   = overall heat transfer coefficient, Btu/hr-ft2-°F
                                                  ATLM = logarithmic mean temperature difference, °F

                                                  and:

                                                  ATLM = [(Tc - The) - (The - Te)] In [(Tc - The)/(The - Te)]

                                                  For dilute emission  streams  that do not require
                                                  additional combustion  air, then ATLM  can be ap-
                                                  proximated by:

                                                  AT, M = Tr -
                                                              he
For a recuperative heat exchanger where the heat
transfer takes place between two gas streams, the
overall heat transfer coefficient (U) ranges from 2
to 8 Btu/hr-ft°F,  generally depending on the heat
exchanger configuration and properties of the gas
streams.

-------
                  Example Case
  Using Equation 4.3-8:
  Qe    =  20,000 scfm
  Me    =  2%
  The    =  550°F (based on heat recovery of 50%)
  Te    =  120T (input data)
  Cpair  =  0.0183 Btu/scf-°F for the interval
           120°-550°F(reference 3)
  U     =  4 Btu/hr-ft2-°F

  Since the emission stream is dilute, calculate
  ATLM as follows:
  AT, M =
  T,
           -Th
   LM —  ' co ' ' he
 co  =  1,000°F
ATLM=  1,000-550 = 450°F
  The heat exchanger surface area from Equation
  4.3-8 then becomes:
        60 x 20,000 x 1.004 x 0.0183 x (550-100)1
  A =
                       4x450
  A =  5,500 ft2

  Using Figure 4-9:
  For all values of he
  Thus, multiplying  Qe by A/Qe, (20,000 x 0.275)
  yields 5,500 ft2.
                     A/Qe is about 275 x 10'
 Alternatively,  Figure 4-9 can be used to determine
 the heat exchanger size. In this figure, line (1) repre-
 sents Equation 4.3-8 evaluated for dilute emission
 streams that do  not require additional  combustion
 air. The assumptions inherent in  this case are the
 same as those stated for Figure 4-7. The overall heat
 transfer coefficient is assumed as 4 Btu/hr-ft2-°F.

 For emission  streams that require additional  com-
 bustion air, use the solid line in Figure 4-9 to obtain
 an estimate of the required heat exchanger size. As
 in Figure 4-8, the solid line is  based on maximum
 combustion air requirements (i.e., no oxygen in the
 emission stream). If Figure 4-9 is not directly appli-
 cable in a particular situation, use the broken line in
 the figure to obtain a conservative estimate for A.

4.3.7 Evaluation  of Permit Application
 Compare the  results from the  calculations and the
 values supplied by the permit applicant using Table
 4-4. The calculated values in the table are based on
 the example case.

 If the calculated  values for Hf, Qc, Qcom, Vbed, and
 A differ from the reported values for these varia-
 bles,  the differences may be  due to the assump-
 tions involved in the calculations.  If that is the case,
 the reviewer may wish to discuss the details of the
 proposed design with the permit applicant.
                                                   If the calculated values agree  with the reported
                                                   values, however, then the design and operation of
                                                   the proposed  catalytic incinerator system may be
                                                   considered appropriate based on the assumptions
                                                   used in this handbook.
                                                   Figure 4-9.  Heat exchanger size vs. emission stream heat
                                                             content.
S 340
§
f 330
1
0)
° 320
CD
CD
£C
310
LL
E
CD
CD
co 300
0
'co
CO
E 290
LU
t- 28°
CD
CD
% 270
t
D
CO
CD
g> 260
CD
.C
O
X
LU
*- 250
m our.
o> 0
I
p 280 1 	
(D
§ " <2) V
0 240 - \
X _ \
S? 200 - \
^. ^L
' ' ~ ^V
\
16A I I 1 I I 1 I I I I | I 1 I
0 2 46 8 10 12 14
he

(1) Dilute stream/no combustion air
(2) No oxygen in emission stream/

maximum combustion air


_
(1)

_




-
(2) _

^

f l I I I
12345

                                                   Table 4-4.
                                                              Emission Stream Heat Content, he (Btu/scf)
Comparison of Calculated Values and Values
Supplied by the Permit Applicant for Catalytic
Incineration
                                                                               Calculated
                                                                                Value
                                                                            (Example Case)"
                              Reported
                                Value
                                                   Supplementary heat
                                                    requirement, Hf
                                                   Supplementary fuel flow
                                                    rate,Qf
                                                   Combustion air flow rate, Qc
                                                   Combined gas stream flow
                                                    rate, Qcom
                                                   Catalyst bed volume, Vbed
                                                   Heat exchanger surface area (if
                                                    recuperative heat recovery
                                                    is used), A
                 150,500 Btu

                  170 scfm
                     0

                 20,170 scfm
                   40ft3
                  5,500 ft2
                                                   a Based on Emission Stream 2.
                                                                           59

-------
4.4  Flares
Flares use open flames for disposing of waste gas-
es during  normal operations and  emergencies.
They are typically applied when the heating value
of the waste gases cannot be recovered economi-
cally because of intermittent or uncertain flow, or
when process upsets occur. In some cases, flares
are operated  in conjunction with baseload gas re-
covery systems  (e.g., condensers).  Flares handle
process upset and emergency gas releases that the
baseload system is not designed to recover.

There are several types of flares, the most common
of which are steam-assisted, air-assisted, and pres-
sure head flares. Typical flare operations can be
classified as  "smokeless," "nonsmokeless,"  and
"fired"  or "endothermic." For  smokeless oper-
ation, flares use  outside momentum sources (usu-
ally steam or  air) to provide efficient gas/air mixing
and  turbulence for complete combustion. Smoke-
less  flaring is required for destruction of organics
heavier than methane.  Nonsmokeless operation is
used for organic or other vapor streams which burn
readily and do not  produce smoke.  Fired, or en-
dothermic, flaring requires additional energy in or-
der  to ensure complete oxidation  of the waste
streams  such as for sulfur tail gas and ammonia
waste streams.

In general, flare performance depends on such fac-
tors  as flare  gas exit velocity,  emission stream
heating value, residence time in  the combustion
zone, waste gas/oxygen mixing,  and flame tem-
perature. Since  steam-assisted smokeless flares
are the most frequently used, they will be the focus
of this discussion. A typical  steam-assisted flare
system is shown in Figure 4-10. First, process off-
gases enter the flare through the collection header.
When water or organic droplets are present, it may
be necessary to pass the off-gases through a knock-
out drum, since these droplets can create prob-
lems. Water droplets can extinguish the flame and
organic droplets can result in burning particles.

Once the off-gases  enter the flare stack, flame
flashback can occur if the emission stream flow
rate is too low. Flashback may be  prevented, how-
ever,  by  passing the gas through a gas barrier, a
water seal, or a stack seal. Purge gas is another
option. At the flare tip, the emission stream is ignit-
ed by pilot burners. If conditions in the flame zone
are optimum (oxygen availability, adequate resi-
dence time,  etc.), the VOC in  the emission stream
may be completely burned (-100  percent efficien-
cy). In some cases,  it may be necessary to add
supplementary fuel (natural gas)  to the emission
stream in order to achieve destruction efficiencies
of 98 percent and greater if the net heating value of
the emission stream is less than 300 Btu/scf.(9,10)

Typically, existing flare systems  will  be used to
control HAP emission streams. Therefore, the fol-
Figure 4-10.  A typical steam-assisted flare system.
                                                  Steam
                                                 Nozzles
                 Gas Collection Header
                   and Transfer Line
         Emission
           Stream'
                  Knock-out Drum
                          Pilot Burners
                           Steam Line

                           Ignition Device


                           Air Line
                           Gas Line
                                   Drain
                      60

-------
lowing sections describe how to evaluate the de-
struction efficiency of an existing flare system un-
der  expected flow conditions (e.g., continuous,
start-up, shut-down, etc.). The discussion will be
based on the recent regulatory requirements of 98
percent destruction efficiency for flares.(9) The cal-
culation procedure will be illustrated for Emission
Stream 3 described  in Chapter 3 using a steam-
assisted flare system. Note that flares often serve
more than one process unit and the total flow rate
to the flare needs to  be determined before the fol-
lowing calculation procedure can  be applied.

4.4.1  Data Required
The  data necessary to perform  the calculations
consist of HAP emission stream characteristics pre-
viously compiled  on the HAP  Emission Stream
Data Form, flare dimensions, and  the required HAP
control as determined by the applicable regula-
tions.
                 Example Case
  Expected emission stream flowrate,
  Qe = 30,000 scfm
  Emission stream temperature, Te = 100°F
  Heat content, he = 180 Btu/scf
  Mean molecular weight of emission stream,
  MWe = 33.5 Ib/lb-mole
  Flare tip diameter, Dtlp = 48 in

  Based on the control requirements for the emis-
  sion stream:
  Required destruction efficiency, DE = 98%
In the case of a permit review, the  data outlined
below should be supplied by  the applicant. The
calculations in this section will then be used to
check the applicant's values. Worksheets are pro-
vided in Appendix C.5.

Flare system variables at standard conditions (70°F,
1 atm):

Flare tip diameter, Dtip, in
Expected emission stream flowrate, Qe, scfm
Emission stream heat content, he, Btu/scf
Temperature of emission stream, Te,  °F
Mean molecular weight of emission stream,
MWe, Ib/lb-mole
Steam flowrate, Qs, Ib/min
Flare gas exit velocity, Uf|g, ft/sec
Supplementary fuel flow rate,3 Qf, scfm
Supplementary fuel heat content,3 hf, Btu/scf
Temperature of flare gas,b Tf|g, °F
Flare gas flow rate,b Qfig, scfm
Flare gas heat content,  hf|g, Btu/scf
4.4.2 Determination of Flare Operating Variables
Based  on studies conducted by EPA, relief gases
having heating values less than 300 Btu/scf are not
assured of achieving 98  percent destruction effi-
ciency when they are flared in steam- or air-assist-
ed flares.0 Therefore, the first step in the evaluation
procedure is to check the heat content of the emis-
sion stream  and determine if additional  fuel is
needed.

In a permit review case, if the heat content of the
emission stream is less than 300 Btu/scf and  no
supplementary fuel has been added, then the appli-
cation  is considered  unacceptable.  The reviewer
may then wish to follow the calculations described
below.  If the  reported  value for the emission
stream heat content is above 300 Btu/scf, however,
then the reviewer should skip to Section 4.4.2.3.

4.4.2.1  Supplementary Fuel Requirements
If the emission stream heat content is less than the
300 Btu/scf required to achieve a  destruction level
of 98 percent, it is assumed that natural gas will be
added to the emission stream to bring its heat con-
tent to 300 Btu/scf. Calculate the required  natural
gas requirements using the following equation:

Qf = [(300 - he)Qe]/582                    (4.4-1)

where:
Qe = emission stream flow rate, scfm
Qf  = natural gas flow rate, scfm
he  = emission stream heat content, Btu/scf

(See Appendix B.7, reference 8, for details of all the
equations used  in Section 4.4.) If  the emission
stream heat content is greater than or equal to 300
Btu/scf, then Qf = 0.
                 Example Case
  Using Equation 4.4-1:
  Since he is less than 300 Btu/scf, supplementary
  fuel is needed.
  he  = 180 Btu/scf
  Qe = 30,000 scfm
  Qf  = (300-180)(30,000)/582
  Qf  = 6,200 scfm
"This information is needed if the emission stream heat content is less
 than 300 Btu/scf.
b If no auxiliary fuel is added, the value for this variable will be the same
 as that for the emission stream.
Tor unassisted flares, the lower limit is 200 Btu/scf.
4.4.2.2 Flare Gas Flow Rate and Heat Content
The flare gas flow rate is determined from the flow
rates of the emission stream and natural gas using
the following equation:

Qflfl = Qe + Qf                           (4.4-2)

where:
Qfig = flare gas flow rate, scfm

Note that if Qf = 0, then Qf,g = Qe.

The heat content of the flare gas (hf|g) is dependent
on whether supplementary fuel is  added to the
                                                                        61

-------
emission stream. When he is greater than or equal
to 300 Btu/scf, then hflg =  he. If he is less than 300
Btu/scf, since supplementary fuel is added  to in-
crease he to 300 Btu/scf, hfig = 300 Btu/scf.
                 Example Case
 Using Equation 4.4-2:
 Qe   = 30,000 scf m
 Qf   = 6,200 scf m
 Qfig  = 36,200 scfm
 Since he =  180 Btu/scf, hf,g = 300 Btu/scf.
4.4.2.3 Flare Gas Exit Velocity
The flare gas exit velocity values presented in Table
4-5 to achieve at least 98 percent destruction effi-
ciency in  a steam-assisted flare system are based
on  studies conducted by EPA.O) Flare gas exit ve-
locities are expressed as a function of flare gas heat
content.  Determine the maximum  allowable exit
velocity using the equation presented in Table 4-5.

Table 4-5.    Flare Gas Exit Velocities (9)
  Flare Gas Heat Content3
      hflg (Btu/scf)
                          Maximum Exit Velocity
                              Umax (ft/sec)
       h,ig < 300
    300 s h,,g < 1,000
      h,,g> 1,000
                         3 28
                                  400
8 If no supplementary fuel is used, hfig = he.
b Based on studies conducted by EPA, waste gases having heat-
 ing values less than 300 Btu/scf are not assured of achieving
 98%  destruction efficiency when they are flared in steam-
 assisted flares.(4)
                 Example Case
 Since hf|g  = 300 Btu/scf, use the equation  in
 Table 4-5 to calculate Umax:
 Umax  = 3.28 [io(0-00118 h"= + °-908)
       = 3 28 f-]o(000118 x 30°
 Umax  = 60 ft/sec
                               °-908)l
From the emission stream data (expected flow rate,
temperature)  and information on flare diameter,
calculate the flare gas exit velocity (Ufig); compare
this value with Umax. Use the following equation to
calculate Uf|g:

Ufig = (Qflg,a/60)[4MDtlp/12)2]
    = (3.06 Qfig,a)/(Dtlp)2                   (4.4-3)

where:
Ufig   = exit velocity of flare gas, ft/sec
Qfig.a   = f'are  9as fl°w rate at actual conditions,
         acfm
Dtip   = flare tip diameter, in

Use the following expression to calculate Qfig,a:
Qfig,a =
460)1/530

    62
                                          (4.4-4)
                                                   If Ufig is less than Umax, then the 98 percent de-
                                                   struction  level can  be achieved. However, if Uflg
                                                   exceeds  Umax, this destruction efficiency  level
                                                   cannot be achieved. This indicates that the existing
                                                   flare  diameter is too small to accommodate lower
                                                   exit velocities for the emission stream under con-
                                                   sideration. Note that at very low flare gas  exit ve-
                                                   locities, flame instability may occur. The minimum
                                                   flare  gas exit velocity for a stable flame is assumed
                                                   as 0.03 ft/sec in this handbook.Cl) Thus, if Ufig  is
                                                   below 0.03 ft/sec, the desired destruction efficiency
                                                   may  not be achieved. In summary, Ufig should fall
                                                   in the range 0.03 ft/sec and Umax for a 98  percent
                                                   destruction efficiency level.

                                                   In a permit review case, if Ufig exceeds Umax, then
                                                   the application is not  acceptable. If Ufig is below
                                                   Umax and exceeds 0.03 ft/sec, then  the proposed
                                                   design is considered acceptable and the reviewer
                                                   may  proceed with the calculations.
                                                  Example Case
                                  Using Equations 4.4-3 and -4:
                                  Qfig   = 36,200 scfm
                                  Tf,g   = 95°F
                                  Qfig.a = [36,200 (95 + 460)]/530
                                  Qfig,a = 37,900 acfm
                                                     Ufig

                                                     Uf,g
                                         = 48 in
                                         = (3.06 x 37,900)/(48)2
                                         = 50 ft/sec
                                                     Since 0.03 ft/sec < Uflg  = 50 ft/sec < Umax  =
                                                     60 ft/sec, the required level of 98% destruction
                                                     efficiency can be achieved under these condi-
                                                     tions.
                                 4.4.2.4 Steam Requirements
                                 Steam requirements for steam-assisted flare oper-
                                 ation depend on the  composition of the flare gas
                                 and the flare-tip design. Typical values range from
                                 0.15 to 0.50 Ib steam/lb flare gas. In this handbook,
                                 the amount of steam required  for 98 percent de-
                                 struction efficiency is assumed as 0.4 Ib steam/lb
                                 flare gas.(2) Use the  following  equation to  deter-
                                 mine steam requirements:

                                 Qs = 1.03 x 10~3 x Qfig x MWf|g             (4.4-5)

                                 where:
                                 Qs     = steam requirement, Ib/min
                                        = (Of x 16.7 + QeMWe)/Qf|g
                                                                     Example Case
                                                     Using Equation 4.4-5:
                                                     Qfig    = 36,200 scfm
                                                     MWfig  = 30.6 Ib/lb-mole
                                                     Qs     = 1.03 x10"3x 36,200x30.6
                                                            = 1,140 Ib/min

-------
4.4.3 Evaluation of Permit Application
Compare the  results from  the  calculated and re-
ported values  using Table 4-6. If the calculated val-
ues of Qf, Ufig, Qfig, and  Qs are different from the
reported values for these variables, the differences
may be due to the assumptions  (e.g., steam to flare
gas ratios, etc.) involved in the calculations. In such
a case, the reviewer may wish to discuss the details
of the proposed system with the permit applicant.

If the calculated values  agree  with  the reported
values,  then the operation of the proposed flare
system  may be considered appropriate based on
the assumptions made in this handbook.
Table 4-6.   Comparison of Calculated Values and Values
          Supplied by the Permit Applicant for Flares
Figure 4-11.  Adsorption isotherms for toluene/activated
          carbon system.(11)
Calculated
Value Reported
(Example Case)3 Value
Supplementary fuel flow rate, Qf
Flare gas exit velocity, Uf,g
Flare gas flow rate, Qfig
Steam flow rate, Qs
6,200 scfm
50 ft/sec
36,200 scfm
1,140lb/min
a Based on Emission Stream 3.


4.5  Boilers/Process Heaters
The application of boilers and/or process heaters as
emission control devices is very site-specific (see
Section 3.2.1.4). The level of detail required in the
calculations for sizing such devices is beyond the
scope of this handbook and thus is not presented.

4.6  Carbon Adsorption
Adsorption is a surface phenomenon whereby hy-
drocarbons  and other compounds are selectively
adsorbed on the surface of such materials as acti-
vated carbon, silica gel, or alumina. Activated car-
bon  is the most widely used adsorbent. The ad-
sorption capacity of an adsorbent for a given  VOC
is often represented by adsorption isotherms that
relate the amount of VOC adsorbed (adsorbate) to
the equilibrium pressure (or concentration) at con-
stant temperature (see Figure 4-11). Typically, the
adsorption capacity increases with the molecular
weight  of the VOC being adsorbed.  In addition,
unsaturated compounds are generally more com-
pletely adsorbed than saturated compounds, and
cyclical compounds are more easily adsorbed  than
linearly  structured  materials. Also, the adsorption
capacity is enhanced by lower operating tempera-
tures and higher concentrations. VOC's character-
ized  by  low vapor pressures  are more easily ad-
sorbed than those with high vapor pressures.

Carbon  adsorption  is used for  pollution  control
and/or for solvent recovery in several industries. It
is  usually a batch  operation, involving  multiple
beds. The two main steps in the adsorption oper-
                       500 |     5,000 |    50,000
                100       1,000      10,000   100,000
               Toluene Concentration (ppmv)
ation include adsorption and regeneration, usually
performed cyclically. For control of continuous
emission streams, at least one bed remains on line
in the  adsorption mode  while the other is being
regenerated.

A typical batch operation (see Figure 4-12) can be
described as follows: The VOC-laden waste gas is
passed through the carbon bed  where the VOC's
are adsorbed on the bed surface. As the adsorption
capacity of the bed is approached, traces of VOC's
appear in the exit stream, indicating that the break-
through point of the bed has been attained. The
emission stream is then directed to a parallel bed
containing regenerated adsorbent, and the process
continued. Concurrently, the  saturated bed is re-
generated by the passage of  hot inert gases, low-
pressure steam, or a combination of vacuum and
hot gas. Since adsorption is a reversible process,
by  supplying heat (equivalent to the amount of
heat released during  adsorption), the VOC's on the
bed can be desorbed. A "heel" is always left on the
bed because complete desorption is technically dif-
ficult to achieve and economically impractical.

During the last part of the steam regeneration cy-
cle, the hot bed saturated with water vapor is dried
and cooled, usually with air. When steam is used as
the regenerant, the desorbed VOC's in the steam
effluent are typically condensed. Then the VOC's
are recovered either  by simple decantation, in the
case of water-insoluble materials, or by distillation,
in the case of water-soluble materials. If high purity
is required for the recovered VOC's, complex distil-
lation  systems may be  necessary,  especially  in
cases where the VOC's consist of mixtures of sol-
vents.

In other designs, continuous adsorption can be ac-
complished by fluidized bed adsorption. The fresh
adsorbent flows down the adsorption section that
consists of a series of fluidized trays. The emission
                                                                       63

-------
Figure 4-12.  A typical fixed-bed carbon adsorption system.


                                  Emission Stream
                                                                           Condenser
      Low-Pressure
        Steam
stream enters at the bottom of the adsorption sec-
tion and flows upward. The VOC's are progressive-
ly adsorbed and the exit gas is discharged from the
top stage. The saturated adsorbent is continuously
removed from the bottom and  transferred to the
desorption section, where it is regenerated and re-
turned to the adsorption system.

For a given emission stream, the performance of a
carbon adsorber as a control device is affected by
several variables including: (a) the adsorption ca-
pacity of the carbon for the specific VOC  in ques-
tion (as determined from the adsorption isotherm),
(b) operating temperature, (c) adsorption and re-
generation cycle time, (d) amount and type of re-
generant, and (e) contaminants.

The discussion in the following sections will be
based on a fixed-bed carbon  adsorption system
with two parallel beds.  Regeneration of the beds
will be carried out with low pressure steam.  It  is
assumed that the desorbed VOC's  and steam  will
be condensed and the bed will be dried and cooled
with air.  Another assumption is that the emission
stream is free from liquid and/or solid particles that
may potentially  blind the carbon beds. Emission
Stream 4 described  in Chapter 3 will be used to
illustrate the calculation procedure. Worksheets for
calculations are provided in Appendix C.6.


4.6.1 Data Required
The  data necessary to  perform the calculations
consist of HAP emission stream  characteristics  pre-
viously  compiled  on the HAP Emission Stream
Data Form and the required HAP control as deter-
mined by the applicable regulations.

                      64
                                                                            Exhaust to
                                                                           Atmosphere
                 Example Case
 Maximum flowrate, Qe = 15,000 scfm
 Temperature, Te = 90°F
 Relative humidity, Rhum = 40%
 HAP = toluene
 Maximum HAP content, HAPe = 1,000 ppmv

 Based on the control requirements for the emis-
 sion stream:
 Required removal efficiency, RE = 95%
If dilution air is added to the emission stream upon
exit from process, the data that will be used in the
calculations are the resulting characteristics after
dilution.

In a permit review case for a carbon adsorber, the
following data should be supplied by the applicant.
The calculations in this section will later be used to
check the applicant's values.

Carbon  adsorber (fixed-bed)  system variables at
standard conditions (70°F, 1  atm):

Reported removal efficiency, RE^oned- °/°
HAP content,  HAPe, ppmv
Emission stream flow rate, Qe, scfm
Adsorption capacity of carbon bed, AC, Ib  HAP/100
Ib carbon
Number of beds, IN
Amount of carbon required, Creq, Ib
Cycle time for adsorption, 6ad, hr
Cycle time for regeneration, 0reg, hr
Emission stream velocity through the carbon bed,
Ue, ft/min
Bed depth, Zbed, ft
Bed diameter, Dbed, ft
Steam ratio, St, Ib steam/lb carbon

-------
4.6.2 Pretreatment of the Emission Stream

4.6.2.1 Cooling
Adsorption of VOC's is favored by  lower tempera-
tures. If the temperature of the emission stream is
significantly higher than  100°F, a heat exchanger
may be  needed to cool  the emission stream to
100°F.  (See Appendix B.5, reference 8, for deter-
mining the size of a heat exchanger required  for
such applications.
content will be limited to less than 25 percent of the
LEL See Table B.1-2  in Appendix B.1 for a list of
LEL values for several compounds.
                 Example Case
 The HAP concentration of the emission stream is
 1,000 ppmv (toluene). This is below 25% of the
 LEL for toluene, which is 3,000 ppmv (see Table
 B.1-2).
                 Example Case
 The temperature of the emission stream is 90°F,
 which is below 100°F. Therefore, cooling  is not
 necessary.
4.6.2.2 Dehumidification
Since water vapor competes with the VOC's in the
emission stream for adsorption sites on the carbon
surface, emission stream humidity levels exceed-
ing 50 percent (relative humidity) are not desirable.
In this handbook, it is assumed that if the relative
humidity level of the emission stream is above 50
percent, it will be reduced to 50 percent using addi-
tional equipment.

Dehumidification may be carried out by cooling
and condensing the water vapor  in the emission
stream. A shell-and-tube type heat exchanger can
be employed for this purpose. Refer to Section 4.8
where calculation procedures for sizing condensers
are described.

Another alternative  for dehumidification  is adding
dilution airto the emission stream if the dilution air
humidity is significantly  less than that of the emis-
sion stream. However, since this will increase the
size of the adsorber  system required, it may not be
cost effective.
                 Example Case
 Since  the  relative  humidity of the emission
 stream is less than 50%, dehumidification is not
 necessary.
4.6.2.3 High VOC Concentrations
If flammable vapors  are  present in emission
streams that are mixtures of VOC and air, the VOC
content may be limited to below 25 percent of the
LEL by insurance companies. In some cases, it can
be increased to 40 to 50 percent of the LEL if proper
monitoring and controls are  used. In addition, since
high bed temperatures may occur due to  heat re-
leased during adsorption, high VOC concentrations
may need to  be  reduced. The maximum practical
inlet VOC concentration is  usually  about 10,000
ppmv. In this handbook, it is assumed that the VOC
4.6.3 Carbon Adsorption System Design Variables
Table 4-7 presents suggested values for the design
variables of a carbon adsorber system to achieve a
given outlet HAP concentration. If the emission lim-
it requirement is expressed  as removal efficiency,
the outlet  HAP concentration can  be calculated
from the required removal efficiency and the inlet
HAP concentration.
Table 4-7.   Carbon Adsorber System Design Variables (12,13)

                                Steam Requirement
    Outlet    Adsorption Degeneration  for Regeneration
Concentration3 Cycle Time Cycle Timeb    St(lb steam/
 HAP0 (ppmv)    M-ad(hr)     (jLreg (hr)       Ib carbon)
     70
    10-12
0.3
1.0
aEmission stream exiting the carbon adsorber.
bRegeneration cycle also includes the time necessary for drying
 and cooling the bed.
For specific applications, other values may be ap-
propriate depending on the emission stream  char-
acteristics and the type of carbon bed. For example,
the adsorption capacity for a given carbon bed  is
dependent on several factors, including the type of
VOC in the emission stream and the temperature
and humidity levels. Typically, the adsorption ca-
pacity is determined from the adsorption isotherm
of the compound under consideration. (See Appen-
dix B.8, reference  8, or references 14 and 15 for
adsorption isotherms for  several compounds. Also
see reference 8 or 16 for adsorption capacities at
specific conditions for several compounds.)

Based  on the required removal  efficiency, deter-
mine the  outlet HAP concentration  using the fol-
lowing equation:
HAP0 = HAPe(1 -0.01 RE)
  (4.6-1)
where:
HAP0 = HAP content of the emission stream exit-
        ing the adsorber, ppmv
RE   = removal efficiency, %

Next, specify the appropriate values for 0ad, 6req,
and St using Table 4-7.
                                                                       65

-------
                 Example Case
 Using Equation 4.6-1:
 RE   = 95 percent
 HAPe = 1,000 ppmv
 HAP0= 1,000(1 -0.95)
 HAP0 = 50 ppmv

 Assuming the conditions for HAP0 = 70 ppmv
 are approximately the same as those for HAP0
  = 50 ppmv, from Table 4-7,
 6ad  = 2 hrs
 ereg  = 2 hrs
 St   = 0.3 Ib steam/lb carbon
4.6.4 Determination of Carbon Adsorber System
     Variables

4.6.4.1  Carbon Requirements
In sizing a carbon adsorber system, the quantity of
carbon required is determined from the adsorption
capacity of the carbon bed (based on the adsorp-
tion isotherm of  the HAP in question) using  the
emission stream flow rate and HAP concentration.
For a fixed-bed adsorption system with N parallel
beds and a specified adsorption cycle 0adr the follow-
ing equation can be used to calculate the carbon
requirements:

Cr
-eq = 2 [1.55x 10"5N6adQe
     (HAPe - HAP0)MWHAp/AC]
                                         (4.6-2)
where:
Weq
N
ead
Qe
MWHAP
AC
       = carbon requirement, Ib carbon
       = number of carbon beds
       = adsorption cycle time, hr
       = emission stream flow rate, scfm
       = molecular  weight of HAP,  Ib/lb-mole
         (for a mixture of HAPs, MWHAp will be
         defined as mean molecular weight)
       = adsorption capacity of the carbon bed,
         Ib HAP/100 Ib carbon
For design purposes, the  carbon  requirement  is
generally multiplied by a factor of 2 as indicated  in
Equation 4.6-2.(17)  This safety factor is an
allowance for build-up of a heel during regenera-
tion (which results in  a reduced capacity); fluctu-
ations in emission stream characteristics (e.g., HAP
concentration  and  composition, humidity,  etc.).
The value for AC is typically determined from the
adsorption isotherm of the specific  HAP/carbon
system. If an isotherm for the HAP in question  is
not available, the isotherm for another compound
of similar molecular weight and  boiling point may
be used as an approximation. See Table  B.1-3  in
Appendix B.1  for  molecular weight  and boiling
point data for several compounds. For additional
data, see reference 18. If no data are available, use
a conservative value of 5 Ibs HAP/100 Ib carbon for
AC.
                                               As an alternative, Figure 4-13 can be used to deter-
                                               mine carbon requirements. The figure is based on
                                               Equation 4.6-2 and evaluated at HAP0 =  10 and 70
                                               ppmv for several inlet concentrations.

                                               If the emission stream contains a mixture of HAP's,
                                               Equation 4.6-2 should be evaluated using appropri-
                                               ate adsorption isotherms for each component and
                                               then summed to determine Creq.
                Example Case
 Using Equation 4.6-2:
 Assume N = 2
         = 2 hrs (from Table 4-7)
         = 15,000 scfm
         = 1,000 ppmv
         = 50 ppmv
                                                Qe
                                                HAP
                                                HAP
 MWHAp  = MWto,uene = 92 Ib/lb mole
 AC      = 20-25 Ib toluene/100 Ib carbon

 This value is estimated from Figure 4-11 where
 adsorption isotherms for toluene are plotted at
 different temperatures. To obtain  a conservative
 estimate for Creq, assume AC = 20:

 Creq = [2 x 1.55 x 10'5 x 2 x 2 x 15,000 (1,000 -
        50) 92/20]
 Creq = 8,100lb

 Using Figure 4-13:
 For HAPe = 1.000 ppmv, HAP0 = 50 ppmv, and
 AC = 20 Ib, Creq/Qe is about 0.55, assuming the
 curve for HAP0 = 70 ppmv is applicable. Thus,

 Creq = 0.55 x 15,000 = 8,250 Ibs
4.6.4.2 Carbon Adsorber Size
The size of the adsorber is determined using a two
step calculation. First, using the actual flow rate of
the emission stream and its velocity, calculate the
bed area. Typically,  velocities  used in industry
range from 50 to 100 ft/min depending on the sys-
tem pressure. At high velocities, the bed pressure
drop becomes too high for standard blowers;  at
lower velocities, the bed becomes too  large and
expensive. A value of 100 ft/min is assumed in this
discussion. Use the following equation to calculate
the required bed area:
                                               Abed = Qe,a/Ue
                                        (4.6-3)
                                               where:
                                               Abed  = bed area, ft2
                                               Qe-a   = emission stream flow rate at actual condi-
                                                       tions, acfm
                                               Ue    = emission stream velocity, ft/min

                                               In this expression, Qe a is determined as follows:

                                               Qe,a = Qe [Te + 460)/530]                 (4.6-4)
                      66

-------
Figure 4-13. Carbon requirement vs. HAP inlet concentration.
                        MWHAp =100lb/lb-mole
                        N      =2
                        ead     = 2 hr
              1,000                 5,000    10,000

            HAP Inlet Concentration, HAPe (ppmv)
where Qe and Te are the flow rate and temperature
of the emission stream. From the bed area, calcu-
late the bed diameter assuming a circular vessel;
use the following expression:

Dbed = 2 [Abed/Tr]°-5 =1.13 (Abed)°-5         (4.6-5)
where:
Dbed = bed diameter, ft

To calculate the  bed depth, determine the volume
occupied by carbon in each bed. Assume a carbon
bed density of pbed, and  use  the following equa-
tion to calculate the volume of carbon (per bed):
V,
 carbon ~~
q/N)/Pb(
                  ed
(4.6-6)
Having calculated  Vcarbon- the bed depth  can be
determined as follows:
 -bed
     = v
        carbo
n/A:
              'bed
(4.6-7)
where Zbed  is the bed  depth, ft.  Hence,  the  re-
quired adsorber size for an adsorption cycle time
ead for obtaining RE percent removal efficiency for
an emission stream with flowrate Qe is: De ft (di-
ameter) by Zbed ft straight side (minimum). Note
                                      that in cases where large flows (> 20,000 scfm) of
                                      off-gases are handled, three or more parallel beds
                                      may be used,  reducing  the  bed size. The  cycle
                                      times for adsorption and regeneration will change
                                      accordingly.
                                                                   Example Case
                                                   Using Equations 4.6-3, -4, -5, -6, and -7:
                                                         = 90°F
                                                         = 15,000 scfm
                                                         = 15,000[(90 + 460)/530]
                                                         = 15,565 acfm
                                       Te
                                       Qe
                                       Qe.a
                                       Qe,a

                                       Ue    = 100ft/min
                                       Abed  - 15,565/100 = 155.7ft2
                                       Dbed  = 1.13x(155.7)°-5 = 14ft

                                       Assume pbed = 30 Ib/ft3
                                       Vcarbon  = (8,100/2)/30 = 135ft3
                                       Zbed    = 135/155.7 = 0.88 = 1 ft
4.6.4.3 Steam Required for Regeneration
Carbon beds may be regenerated  by various
means; the  most common  regenerant used  is
steam. In this handbook, regeneration with steam
is followed by condensation. The quantity of steam
required for regeneration depends on the required
removal efficiency (or outlet concentration) and on
how much  material is to be desorbed from the bed.
A certain amount of steam is  required to raise the
bed  to its  regeneration temperature and provide
the heat of desorption. The major portion  of the
steam flow, about 60 to 70 percent, acts as a carrier
gas for the desorbed VOC's. It is not cost-effective
to achieve  complete desorption; acceptable work-
ing capacities of adsorption can be obtained with-
out consuming large quantities of steam. For sol-
vent recovery systems a requirement of 0.25 to 0.35
Ib steam/lb carbon has usually been specified. For
applications  where VOC  concentrations are low
(e.g., odor control), steam  usage ratios are higher.

In this handbook, it is assumed that with a steam
ratio of  0.3 Ib steam/lb carbon, a HAP outlet con-
centration of 70 ppmv can  be achieved after regen-
eration, and with a ratio of 1 Ib steam/lb carbon, a
HAP outlet concentration  of  10-12 ppmv  can be
achieved (Table 4-7).(11)  The regeneration  cycle
time, 0reg,  is dependent on the time  required to
regenerate, dry, and cool the bed. The flow rate of
the steam used for regeneration can be determined
using the following expression:
                                                  Qs = [St(creq)/(ereg - edry_coo,)]/60
                                                                              (4.6-8)
                                      where:
                                      Qs       = steam flowrate, Ib/min
                                               = cycle time for drying  and cooling the
                                                 bed, hr
                                                                       67

-------
Typically, cooling and drying the bed with air can
be carried out in about 15 minutes. Figure 4-14 can
also be used to estimate steam requirements. This
figure is based  on 9reg = 2 hrs and  edry.cooi =
0.25 hrs.
Steam flow rates based on cross-sectional area of
the bed (Qs/Abed) are generally limited to less than
4 Ib steam/min-ft2 to prevent the carbon from being
fluidized in the bed. If Qs/Abeci  exceeds 4, the re-
generation cycle time or the steam ratio may need
to be modified.
                 Example Case
  Using Equation 4.6-8:
  RE   = 95%
  HAP0 = 50 ppmv
  St    = 0.3 Ib steam/lb carbon (Table 4-7)
  reg
       = 2 hrs (Table 4-7)
 Assuming edry-cooi = 0.25 hrs:
 Qs      = [0.3 (8,100)7(2 -0.25)1/60
 Qs      = 23 Ib/min
       d  = 23/155.7 = 0.15 I b steam/min-ft2
  Since  Qs/Abed is less than 4 Ib steam/min ft2,
  fluidization in the carbon bed is not expected.

  Using Figure 4-14:
  At Creq = 8,100 Ib and St = 0.3,
  Qs  = 20 Ib/min
Figure 4-14.  Steam requirement vs. carbon requirement.
      100
                  1,000         10,000

               Carbon Requirement, Creq (Ib)
4.6.4.4 Condenser
The steam used for regenerating the carbon bed
containing the desorbed VOC's is  typically con-
densed. The heat duty of the condenser is based on
the amount of steam required to  regenerate the
bed; the amount of heat absorbed by the bed and
later removed by the drying and cooling air can be
considered negligible.

Use the following expression to calculate the con-
denser size:
                                                  ACon - H|0ad/UATLM
                                         (4.6-9)
where:
ACon  = condenser surface area, ft2
      = condenser heat load, Btu/hr
      = overall heat transfer coefficient, Btu/hr-ft2-
        °F
ATLM  = logarithmic mean temperature difference,
        °F
U
and:

ATLM
-[
          (Tsti   Two)   (Tsto   TW|)
          In [(Tsti - Two)/(Tsto - Twi)]
where:
Tsti  = steam inlet temperature, °F
Tsto  = condensed steam outlet temperature,
 I wo
 1 wi
             	    _	   ___-	  p_	
     = cooling water outlet temperature, °F
     = cooling water inlet temperature, °F
In this handbook, it is assumed that following re-
generation,  steam  will  be  condensed  and  sub-
cooled to 100°F with cooling water. Note that Equa-
tion  4.6-9 is an  approximate expression;
condensation and subcooling processes are com-
bined and average values are used for U and ATLM-
To calculate H|oad, use the following equation:

H,oad  = 1.1 x 60 x Qs [X + CPw (Tstl - T^o)]  (4.6-10)
where:
X    = latent heat of vaporization, Btu/lb
Cpw = average specific  heat of water, Btu/lb-°F

In this expression, the condenser heat load is over-
sized  by 10 percent. In the following calculations,
steam available for  regeneration is at atmospheric
pressure and 212°F. The latent heat of vaporization
of steam at these conditions is 970 Btu/lb.(19) The
temperature rise of the  cooling water available at
80°F is 50°F. The overall heat transfer coefficient is
assumed as 150 Btu/hr-ft2-°F.(12)

To determine cooling  water requirements, use the
following equation:

Qcooi.w = H,oad/[Cpw (Two - Twi)]            (4.6-11)

where:
    ,i,w = cooling water flow rate, Ib/hr
    ,i,w can be expressed in terms of gal/min as
follows:
                      68

-------
Qw = QCOOI.W x [(1/60) x (1/62.43) x 7.48]
    = 0.002 X QCOol,w
                               (4.6-12)
where the factor 62.43 is the density of water and
the factor 7.48 is used for converting from "ft3" to
"gal" basis.
                Example Case
Using Equation 4.6-10:
Qs    = 23 Ib/min
X     = 970 Btu/lb (reference 19)
      = 1 Btu/lb-°F (reference 19)
      = 212°F
      = 100°F
      = 1.1 x 60 x 23 [970 + 1 x (212 - 100)]
      = 1,642,500 Btu/hr

Using Equation 4.6-9:
U
 Ts,i
 Tsto
 Hload
 ATLM

 ATLM
= 150Btu/hr-ft2-°F
= [ (212-130)-(100-80)  1
  Lln[(212)130)/(100)80)] J
= 44°F

= 1,642,500/( 150x44)
= 250 ft2
  Using Equations 4.6-11 and -12:
  AT = 50°F
  Qw = 0.002 Qcool.w
  Qw = 0.002 [1,642,500/11x50)1
  Qw = 66 gal/min
4.6.4.5 Recovered Product
To calculate costs, the quantity of recovered prod-
uct that can be sold and/or recycled to the process
has to be calculated. Use the following equation:

Qrec  = 60 x [Qe x (HAPe x 10~6)(1/387)
       (0.01 RE) MWHAp]                 (4.6-13)
Qrec  =  1.55 x  10~9Qex HAPe x RE xMWHAp
where Qrec  is  the quantity of recovered product,
Ib/hr. In this equation, the factor 387 is the volume
(scf) occupied  by 1 Ib-mole of ideal gas at standard
conditions (70°F and 1 atm).
                 Example Case
  Using Equation 4.6-13:
  Qe
  HAPe
  RE
  MWHAp
  Qrec
 = 15,000 scf m
 = 1,000ppmv
 = 95 %
 = 92 Ib/lb-mole
 = 1.55 x10"9x 15,000x1,000x95x92
 = 200 Ib/hr
4.6.5 Evaluation of Permit Application
Compare the results from the calculated values and
reported values using Table 4-8.  If the calculated
Table 4-8.   Comparison of Calculated Values and Values
          Supplied by the Permit Applicant for Carbon
          Adsorption
Calculated
Value Reported
(Example Case)3 Value
Carbon requirement, Creq
Bed diameter, Dbed
Bed depth, Zbed
Steam rate, Qs
Condenser surface area, Acon
Cooling water rate, Qw
Recovered product, Qrec
8,100 Ib
14ft
1 ft
23 Ib/min
250ft2
66gal/min
200 Ib/hr
                                                   'Based on Emission Stream 4.
                                         values of Crpn, Dh
                               Qw, and Qrec are
          'req» "-"bed' ^bed/ QS- ^
different from the reported values for these varia-
bles, the differences may be due to the assump-
tions  involved in the calculations. Therefore, the
reviewer may wish to discuss the  details of the
proposed design with the permit applicant.

If the calculated values agree with  the reported
values, then the design and operation of the pro-
posed carbon adsorber system may be considered
appropriate based on the assumptions made in this
handbook.

4.7 Absorption
Absorption is  an operation in  which one or  more
components of a gas mixture are selectively trans-
ferred into a  relatively nonvolatile  liquid. Absorp-
tion of a gaseous component by a  liquid occurs
when the liquid contains less than the equilibrium
concentration  of the gaseous component. The dif-
ference between the actual concentration and the
equilibrium concentration  provides the driving
force  for absorption. The absorption rate depends
on the physical properties  of the gaseous/liquid
system (e.g., diffusivity, viscosity, density) and the
absorber operating conditions (e.g., temperature,
flow rates of the gaseous and liquid streams). It is
enhanced by lower temperatures, greater contact-
ing surface, higher liquid-gas ratios, and higher
concentrations in the gas stream.

Absorption can be  physical or chemical. Physical
absorption occurs when the absorbed compound
simply dissolves  in the solvent. When there is a
reaction between the absorbed compound and the
solvent, it is termed chemical  absorption. Liquids
commonly used as solvents for organic and inor-
ganic compounds include water, mineral oils, non-
volatile hydrocarbon oils, and aqueous solutions
(e.g.,  sodium hydroxide).

The types of equipment commonly used for gas/
liquid contact operations  include packed towers,
plate  or tray towers, spray chambers, and venturi
scrubbers. These devices are designed to provide
maximum  contact between the gas and  liquid
                                                                       69

-------
streams in order to increase the mass transfer rate
between the two phases. A packed tower is filled
with packing material that is designed to expose a
large wetted surface area to the gas stream. Plate
towers use plates or trays that are arranged so that
the gas stream is dispersed through a layer of liq-
uid on  each plate.  Bubble-cap plates have been
widely used; other types of plates include perforat-
ed trays and valve trays. In  a spray tower, the gas
mixture is contacted with a liquid spray. In a venturi
scrubber, the gas and liquid streams come into
contact at the throat of the venturi nozzle;  venturi
scrubbers are typically used for removal of particu-
late matter (see Section 4.11).

Several different configurations of absorber sys-
tems are used for controlling vapor emissions. The
simplest configuration is one  in which the  solvent
(usually water)  is used on  a  once-through basis,
and is then either discharged to a wastewater treat-
ment system or introduced  as a process water
stream  (see Figure 4-15). The possibility of using
solvents other than water on a once-through basis
may exist when fresh solvent is available in  large
quantities as a  process raw material or fuel. An-
other configuration involves using the solvent
(usually water) on a once-through basis  and strip-
ping it (reverse of absorption) before discharging.
In yet  another configuration, an organic liquid is
used as a solvent and recycled to the absorber after
being stripped.

The efficiency of absorption  for  removing pollu-
tants from gaseous streams  depends on  several
factors, including (a) solubility of the pollutant in a
given solvent, (b) concentration,  (c)  temperature,
 Figure 4-15.  A typical countercurrent packed column absorber
           system.
                            Emission
                            Stream Outlet
       Solvent Inlet
   Emission Stream
            Inlet
                   tTTTt
(d) flow rates of gaseous and liquid streams (liquid
to gas ratio), (e) contact surface area, and (f) effi-
ciency of stripping (if solvent  is recycled to the
absorber).

Determination of the absorber  system variables
(absorber column diameter, height, etc.) is depen-
dent on the individual vapor/liquid equilibrium re-
lationship for the specific HAP/solvent system and
the type of absorber to be used (packed or plate
tower, etc.). Note that equilibrium data may not be
readily available for uncommon HAP's.

Detailed design procedures for all types of absorb-
ers are not appropriate for this handbook; there-
fore, important design considerations for one type
of absorber will be briefly discussed.  Since packed
towers are commonly used in air pollution control,
the discussion will be based on packed tower ab-
sorbers. For illustration purposes, a simple con-
figuration  is chosen for the  absorber system: a
packed tower absorber using 2-inch  ceramic Ras-
chig rings as the packing material with water used
as the absorbent on a once-through  basis. The ef-
fluent from the absorber is  assumed to be dis-
charged to a wastewater treatment facility. The
treatment in the following subsections is equally
applicable to both organic and  inorganic vapor
emissions control. (For more  information on gas
absorption, see references 18, 20, 21, 22, and 23).

As  indicated in Chapter 3, absorption is the most
widely  used control method for inorganic vapor
emissions; therefore. Emission Stream  5 contain-
ing inorganic vapors will be used  in the example
case to illustrate the calculation procedures. Work-
sheets for calculations are provided in Appendix
C.7.

4.7.1  Data Required
The data  necessary  to perform the calculations
consist of HAP emission stream characteristics pre-
viously compiled  on the  HAP Emission Stream
Data  Form and the required HAP control as deter-
mined by the applicable regulations.
                 Example Case
  Maximum flow rate, Qe = 3,000 scfm
  Temperature, Te = 85°F
  HAP = ammonia
  HAP concentration, HAPe = 20,000 ppmv
  Pressure, Pe = 760 mm Hg

  Based on the control requirements for the emis-
  sion stream:
  Required removal efficiency, RE = 98%
                                 Solvent Outlet
 In the case of a permit review for an absorber, the
 following data outlined below should be supplied
 by the  applicant. The calculations  in this section
 will then be used to check the applicant's values.
                      70

-------
Absorption system variables  at standard  condi-
tions (70°F, 1 atm):

Reported removal efficiency, RErep0rted, %
Emission stream flow rate, Qe,  scfm
Temperature of emission stream, Te, °F
HAP
HAP concentration, HAPe, ppmv
Solvent used
Slope of the equilibrium curve, m
Solvent flow rate, Lgai, gal/min
Density of the emission stream, pG, Ib/ft3
Schmidt No. for the HAP/emission  stream  and
  HAP/solvent systems,  ScG, ScL (To calculate ScG
  or ScL, see references  18 or 23 for viscosity, den-
  sity, and diffusivity data.)
Properties of the solvent:
  Density, pL, Ib/ft3
  Viscosity, JJ,L, centipoise
Type of packing used
Packing constants a, b, c, d, e, Y, s, g, r
Column diameter, DCO|Umn, ft
Tower height (packed), HtCO|Umru ft
Pressure drop, APtotai, in H20

4.7.2. Absorption System Design Variables
In absorption, the removal efficiencies  (or outlet
concentrations) are  limited by the driving force
available from gas to the liquid phase. The driving
force for a given set of operating conditions is de-
termined by the difference between the actual  HAP
concentrations in the gas stream and solvent and
the corresponding  equilibrium  concentrations.

When  the slope (m)  of the equilibrium curve  is
small for a given  HAP/solvent system,  indicating
that the HAP is readily soluble in the solvent, the
driving force for absorption is large and absorption
occurs readily. On the other hand, if m is large (e.g.,
> 50), the HAP is not readily soluble in the solvent
and the driving force for absorption is small; there-
fore, long contact times, tall  absorption towers,
and/or high  liquid-gas ratios are required for  ade-
quate performance (high removal efficiency and/or
low outlet concentrations). Hence, as a  conserva-
tive guideline, assume  that if m  is  greater than
about 50 for a given HAP/solvent system at atmo-
spheric pressure, then  high  removal efficiencies
(—99 percent) are not possible.
4.7.3 Determination of Absorber System Design
     and Operating Variables
In most applications involving the  absorption of a
gaseous pollutant from an effluent  gas stream, the
inlet conditions (flow rate,  composition, and tem-
perature) are usually known. The  composition of
the outlet gas is specified by  the control require-
ments. The conditions of the  inlet liquid are  also
known. The main objectives, then, in the design of
an absorption column will be the determination of
the solvent  flow  rate and the calculation  of the
principal dimensions of the equipment (column di-
ameter  and height to  accomplish the  absorption
operation) for a selected solvent.

To keep the  discussion simple, the following as-
sumptions  are made: (1) there are no heat effects
associated  with the absorption  operation, and (2)
both the gas and liquid streams are dilute solutions
(i.e., flow rates are constant throughout the absorp-
tion column and the equilibrium curve can be ap-
proximated as a straight line). All  of the data (e.g.,
packing factors, Schmidt numbers, etc.) required in
the calculation of the design variables can be found
in references 12, 18, 21,  23, 24, or Appendix B.9,
reference 8.

4.7.3.1 Solvent Flow Rate
The quantity  of solvent to be used is typically esti-
mated from the minimum liquid-gas ratio as deter-
mined from material balances and equilibrium con-
siderations. As a rule of thumb  for purposes  of
rapid estimates, it has frequently  been found that
the most economical value for the absorption fac-
tor (defined below) will be in the range from 1.25 to
2.0.{20)
AF =  Lmo,/m G
                                     (4.7-1]
where:
AF    = absorption factor
Lmoi  = liquid (solvent) flow rate, Ib-moles/hr
Gmoi  = gas stream flow  rate, Ib-moles/hr
m    = slope of the equilibrium curve

The value of m is determined from the equilibrium
data at a specific temperature level for the HAP/sol-
vent system under consideration. (See references
12,  21 and 24 for equilibrium for specific systems.
For information on other systems, see references
18,  22, and 23). Assuming a value of 1.6 for AF, use
Equation 4.7-1 to calculate the solvent flow rate:
     - 1.6 m Gm0|
                                     (4.7-2)
The variable Gmoi can be expressed in terms of Qe
as follows:
Gmo, = 0.155 Qe
                                     (4.7-3)
Note that Lmo| can be converted to gal/hr basis as
follows:

Lgai = [Lmo, x MWsolvent x (1/PL) x 7.48]/60    (4.7-4)
where:
Lgai        = solvent flow rate, gal/min
MWSO|vent  = molecular weight of solvent,  Ib/lb-
             mole
PL         = density of solvent (liquid), Ib/ft3

The factor 7.48 is  used to convert from ft3 to gal
basis. For water as the solvent, pL  = 62.43 Ib/ft3
(reference 18) and  MWSO|Vent = 18 Ib/lb mole; then:
= 0.036 Lmo,
                                         (4.7-5)
                                                                        77

-------
                 Example Case
 Using Equations 4.7-2, -3, and -5:
 m    =1.3
 (for the operating conditions in the system, con-
 sult reference 21)
 Qe   = 3,000 scf m
    3l  —
   mol
  -mol
 Lgal
 Lgal
  0.155x3,000x60
  465 Ib-moles/hr

  1.6x1.3x465
  970 Ib-moles/hr

  0.036 x 970
  35 gal/min
4.7.3.2 Column Diameter
Once the gas and liquid streams entering and leav-
ing the absorber column and their concentrations
are identified, flow rates calculated, and operating
conditions (type of packing) determined, the phys-
ical dimensions of the  column can be  calculated.
The column must be of sufficient diameter to ac-
commodate the gas and liquid streams.

The calculation of the column diameter is based on
flooding considerations, the usual operating range
being taken as 60 to 75 percent of the flooding  rate.
One of the commonly  used  correlations in deter-
mining the column diameter is shown in Figure 4-
16. (12) The procedure to calculate the  column di-
ameter is as follows: First, calculate the abscissa
(ABS):

ABS = (L/G) (pc/pj0-5                     (4.7-6)

where:
L   = solvent flow rate, Ib/hr
G   = gas stream flow rate, Ib/hr
pG  = density of emission stream, Ib/ft3

The values for the variables L and G can be calcu-
lated  by  multiplying  Lmo,  and Gmo| with their re-
spective  molecular weights. Then proceed to the
flooding  line in  Figure 4-16 and read the ordinate
(ORD), and solve the ordinate expression for Garea;f
at flooding:
ORD =  [(Garea,f)2 (a/e3)

Thus,
                 °-2)} /pGpLgc
       = {[ORD PG pLgc]/[(a/e3)
                                          (4.7-7)
                                  (4.7-8)
where:
Garea.f
= gas stream flow rate based on column
  cross sectional area (at flooding condi-
  tions), Ib/ft2-sec
= packing factors (see reference 21)
= viscosity of solvent, centipoises
= gravitational constant, ft/sec2
a,e
IJLL
gc

Assuming f as the fraction of flooding velocity ap-
propriate for the  proposed operation, the gas
                                            Figure 4-16.  Correlation for flooding rate in randomly packed
                                                      towers.(12)
                                                i.o
                                                        0.1
                                                   ^  0.01
                                            o
                                               0.001
                                                  0.01    0.045  0.1           1.0

                                                                  (L/G)pG/pL)05
                                                                                        10.0
                Example Case
Using Equations 4.7-6, -7, -8, -9, -10, and-11:
L    =  MWso,vent x Lmol = 18x970
     =  17,460 Ib/hr
G    =  MWe x Gmo,  = 28.4 x 465 = 13,200 Ib/hr
        (see Appendix B.1, reference 8, for
        calculating MWe)
PG   =  0.071 Ib/ft3 (from ideal gas law at 85°F)
        (see reference 21 for calculating pG)
PL   =  62.18 Ib/ft3 (reference 18; at 85°F)
ABS =  (17,460/13,200) (0.071/62.18)° 5 = 0.045
From Figure 4-16, at ABS = 0.045, the  value of
ORD at  flooding conditions is about 0.15. For 2-
inch ceramic Raschig rings, from reference 21:
a = 28
e = 0.74

Also,
gc  = 32.2 ft/sec2
IXL  = 0.85 cp (reference 18, at 85°F)

Thus,
Garea,f =    (0.15x0.071 x 62.18 x 32.2) °-5
           L      [28/(0.74)3] (0.85)0'2   J
     ,,f = 0.56 Ib/sec-ft2 (at flooding)
                                                    Assuming f = 0.60
                                                    Garea = 0.60 x 0.56 = 0.34 Ib/sec-ft2

                                                    Thus,
                                                    ACoiumn   =13,200/(3,600x0.34)
                                                    AColumn   =  10.8ft2
                                                    Dcolumn   =  1.13(10.8)°'5 = 3.7~ 4 ft
                       72

-------
stream flow rate (based  on cross-sectional  area)
can be expressed as:

Garea = f Garea,f                          (4.7-9)

The  usual column operating range for f is taken as
0.60 to 0.75. Calculate the column cross-sectional
area by the following expression:

Aco,umn = g/(3,600 Garea)                  (4.7-10)

The column diameter is then determined by:

Dcoiumn = (4M(Acolumn)°-5  = 1.13(Acolumn)°'5 (4.7-11)

where:
        = column diameter, ft
Figure 4-17. NOG for absorption columns with constant
          absorption factor AF.(12)
4.7.3.3 Column Height
The column must be of sufficient height to ensure
that the required removal efficiency is achieved.
The height of a  packed column is calculated by
determining  the  required number  of  theoretical
transfer  units and multiplying by the height of a
transfer  unit.  A transfer unit  is a measure of the
difficulty of the mass transfer operation and is a
function of the solubility and concentrations of the
solute in the gas and liquid streams. It is expressed
as NOG or  NOL depending on whether the gas film
or  liquid film  resistance controls the  absorption
rate. In emission control applications, gas film re-
sistance will typically be controlling, therefore N0c
will be used in the following calculations.

The expression for the column height (packed) is:

Htcoiumn  =  NOG HOG                       (4.7-12)
where:
HtCoiumn    =  packed column height, ft
NOG       =  number of gas transfer units (based
             on overall  gas film coefficients)
HOG       =  height  of an overall gas transfer unit
             (based on overall  gas film coeffi-
             cients), ft

Although the determination of NOG  is usually com-
plicated, when dilute solutions are  involved, NOG
can be calculated using  the following equation:

NOG  =  In  {(HAPe/HAP0)[1 - (1/AF)] + (1/AF)}/
         [1-11/AF)]                       (4.7-13)
This expression is simplified based on the assump-
tion that no  HAP is present  in the solvent as it
enters the  column (see  reference 8  for details). Al-
ternatively, use Figure  4-17 directly to determine
NOG-
The variable  H0c is  generally calculated from the
following equation:

HOG = HG  + (1/AF) HL                   (4.7-14)

where:
HG   = height of a gas transfer unit, ft
HL   = height of a liquid  transfer unit, ft
                         100

                        HAPe/HAPn
                                  1,000
                                          10,000
Generalized correlations are available to calculate
HG and HL; these are based on the type of packing
and the gas and solvent flow rates. The correlations
for HG and HL are as follows:(12)

                                        (4.7-15)

                                        (4.7-16)
                               05
HL  = Y(L"VL")S (ScL>° 5
HG  = [b (3,600 Garea)c/(L")dl (ScG)

                 =L
where:
b, c, d, Y, and s = empirical packing constants (see
                 reference 18)
L"             = liquid flow rate, Ib/hr-ft2
JAL"            = liquid viscosity, Ib/ft-hr
ScG            = Schmidt number for the gas
                 stream
ScL            = Schmidt number for the liquid
                 stream

The values for ScG and ScL are listed for  several
compounds in references  12 and 23 (or see Appen-
dix B.9, reference 8). In the calculations, it is  as-
sumed that the effect of temperature on Sc is negli-
gible. The value for the variable L" in this equation
is calculated as follows:

L" = L/Acolumn                            (4.7-17)

Use the following expression to calculate the total
column  height (Ht,otai):(24)

Ht,otai = Htco,umn +  2  + 0.25 Dco,umn       (4.7-18)
                                                                         73

-------
                 Example Case
 1. Calculation of NOG
    Using Equation 4.7-13:
    HAPe = 20,000 ppmv
    HAP0 = 20,000 (1 - 0.98)  = 400 ppmv
    NOG  = ln[(20,000/400) 0.375 + 0.625]/0.375
    NOG  = 7.9
    Using Figure 4-17:
    HAPe/HAP0 = 20,000/400 = 50
    At AF = 1.6,  1/AF = 0.63, and NOG = 8

 2. Calculation of HOG:
    Using Equations 4.7-14, -15, -16, and -17:
    L" = 17,460/10.8 = 1,617 Ib/hr-ft2
    3,600 Garea = 1,224 Ib/hr-ft2

    From reference 12, the packing factors are:
    b = 3.82
    c = 0.41
    d = 0.45
    Y= 0.0125
    s = 0.22

    Although  1,224 Ib/hr-ft2 is outside the range
    shown in  the table, assume that the packing
    factors  are applicable and  the error intro-
    duced into the calculations will be negligible.
    From reference 12:
    ScG  = 0.66
    ScL  = 570

    Also,
    M.L" = 0.85 x  2.42 = 2.06  Ib/ft-hr
    (The factor 2.42 is used to convert from centi-
    poise to Ib/ft-hr.)

    Hence,
    HG   =  [3.82(1,224)a41/(1,617)°-45](0.66)a5
         =  2.06
    HL   =  0.0125 (1,617/2.06)022(570)05
         =  1.29
    Using AF  =  1.6,
    HOG =  2.06  +  (1/1.6) 1.29 =  2.87 -2.9
 3. Calculation of Htcolumn:
    Using Equation 4.7-12:
            = 7.9 x 2.9 = 22.9 -23 ft
  4. Calculation of Httotai:
    Using Equation 4.7-18:
    Httotai = 23 + 2 + (0.25 x 4) = 26 ft
For costing purposes, it is necessary to calculate
the column weight. Use the following equation:(24)

Wtcolumn = (48 Dco|umn X Httota|) + 39 (Dco|umn)
                                         (4.7-19)
where:
WtCoiumn = column weight, Ib

Also, to determine packing costs, volume occupied
by the packing material (Vpacking) has to be calculated.
Use the following expression:

^packing = '/1T'4)(lJcolumn'  X Htco|umn
"column ~~ "•'**•'>'•'column'  ^ "^column
                           (4.7-20)
                 Example Case
 Using Equation 4.7-19:
 '-'column    — 4 ft
 Httotai     = 26 ft
 WtCO|umn   = (48 x 4 x 26) + 39(4)2
 Wtco,umn   = 5,600 Ib

 Using Equation 4.7-20:
 HtColumn  = 23 ft
 Vpacking   = 0.785 X (4)2 X 23
 Vpacking   = 290 ft3
4.7.3.4 Pressure Drop Through the Column
The pressure drop through  a packed column for
any combination  of  liquid and gas flows in the
operable range is an important economic consider-
ation in the design of such columns. For a particu-
lar  packing, the most accurate data will be those
available from the manufacturer.  For  purposes of
estimation, use the following correlation:(12)

APa = (g x 10-8) [10(r^L)] (3,600 Garea)2/pG  (4.7-21)

where:
APa = pressure drop, Ib/ft2-ft
g, r = packing constants (see reference 12)

The total pressure drop through the  column is then
expressed as:
             X Ht
                 column
                            (4.7-22)
                  Example Case
 Using Equation 4.7-21 :
 From reference 12:
 g = 11.13
 r  = 0.00295
Also,
L"
3,600 Garea
pG
PL

Thus,
=1,617 Ib/hr-ft2
= 3,600 x 0.34 =
= 0.071 Ib/ft3
=62. 18 Ib/ft3
                              1,224 Ib/hr-ft2
                                                                1Q-8x 10(0-00295X1-617/62'18>(1,224)2]
  APa = 2.8 Ib/ft2-ft
                     (0.071)
  Using Equation 4.7-22:
  Htcolumn   = 23 ft
  AP.otai     = 2.8 x 23 = 64.4 Ib/ft2
  AP,otai     = 64.4/5.2 = 12inH2O
  (The factor 5.2 is used to convert from Ib/ft2 to in
  H20.)
                       74

-------
4.7.4 Evaluation of Permit Application
Compare the results from the calculations and the
values supplied by the permit applicant using Table
4-9. The calculated values in the table  are  based
on the Example Case.  If the calculated values of
Lgaii  Dcoiumn/ HtCO|Umn, Httota|, APtota| Wtco|umn, and
Vpacking are different from the reported  values for
these variables, the differences may be  due  to the
assumptions involved  in the calculations. There-
fore, the reviewer may wish to discuss the details
of the proposed design with the permit applicant.

If the calculated values agree with the reported
values, then the design and operation of the pro-
posed scrubber system may be considered appro-
priate based on  the  assumptions made in this
handbook.
Table 4-9.   Comparison of Calculated Values and Values
          Supplied by the Permit Applicant for Absorption
Calculated
Value Reported
(Example Case)3 Value
Solvent flow rate, Lgai
Column diameter, Dco,umn
Column height, HtCO|umn
Total column height, Httotai
Packing volume, Vpacking
Pressure drop, APtotai
Column weight, WtCO|umn
35gal/min
4ft
23ft
26ft
290 ft3
12inH20
5,600 Ib
aBased on Emission Stream 5.
4.8 Condensation
Condensation is a separation technique in which
one or more volatile components of a vapor mix-
ture  are separated from  the remaining vapors
through saturation followed  by  a  phase change
(see Figure 4-18). The phase  change from gas to
liquid can be accomplished in two ways:  (a) the
system pressure may be increased at a given tem-
perature, or (b)  the system temperature may be
reduced at constant pressure.
The design and operation of a condenser is signifi-
cantly affected by the number and  nature of the
components present in the emission stream. In a
two-component  vapor system where  one of the
components is noncondensible (e.g., air), conden-
sation occurs at dew point (saturation) when the
partial pressure  of the condensible  compound
(e.g., benzene)  is equal  to its vapor pressure. In
most  HAP control applications,  the emission
stream will contain large quantities of nonconden-
sible and small quantities of condensible com-
pounds. To separate the condensible  component
from the gas stream at a fixed pressure, the tem-
perature of the  gas stream must be reduced. The
more volatile a compound (i.e., the lower the nor-
mal boiling point), the larger the amount that can
Figure 4-18. Flow diagram for a typical condensation system
          with refrigeration.
                 Emission Stream Outlet
                	t
Emission
 Stream
   Inlet
Condenser
                                  Condensed VOC
                 Coolant '
                       Refrigeration
                          Unit
remain as vapor at a given temperature; hence the
lower the temperature required for saturation (con-
densation).

When condensers are used to control emissions,
they are usually operated at the constant pressure
of the emission source, which is normally close to
atmospheric. Depending  on  the temperatures re-
quired for condensation, a refrigeration  unit may
be necessary to supply the  coolant  (see Section
4.8.3.2). The two most common types of condens-
ers used are surface and contact condensers. Sur-
face condensers are usually shell-and-tube heat ex-
changers.  The coolant typically flows through the
tubes and the vapors condense on the shell outside
the tubes. The condensed vapor forms a film on the
cool  tubes and is drained to a collection tank for
storage or disposal. In contrast to surface condens-
ers where the coolant does not contact either the
vapors or the condensate, in contact condensers,
the vapor mixture  is cooled by spraying a cool
liquid directly into the gas stream.

Design calculations for condenser systems vary in
complexity depending on the nature and number
of components present in the emission stream. For
detailed information on condenser design, consult
references 25 and 26. In the  following discussion,
Emission Stream 6, consisting of a single condens-
ible component and a single noncondensible com-
ponent, will be used to  illustrate the calculation
procedure for surface condensers.  It will be as-
sumed that the moisture content  of the emission
stream is negligible (i.e., no ice is expected to form
on the tubes in the condenser). The design proce-
dure will  involve  determining  the  condensation
temperature required, selection of coolant, and cal-
culation of condenser size and coolant require-
ments.

4.8.1. Data Required
The  data  necessary to  perform the calculations
consist of HAP emission stream characteristics pre-
                                                                       75

-------
viously  compiled on  the  HAP Emission Stream
Data Form and the required HAP control as deter-
mined by the applicable regulations.
                 Example Case
 Maximum flow rate, Qe = 2,000 scfm
 Temperature, Te = 90°F
 HAP = styrene
 HAP concentration, HAPe = 13,000 ppmv
   (corresponding to saturation conditions)
 Moisture content, Me = negligible
 Pressure, Pe = 760 mm Hg

 Based on the control requirements for the emis-
 sion stream:
 Required removal efficiency, RE = 90%
In the case of a permit review for a condenser, the
following data should be supplied by the applicant.
The calculations in this section will then be used to
check the applicant's values. Worksheets for calcu-
lations are provided in Appendix C.8.

Condenser system variables at standard conditions
(70°F, 1 atm):

Reported removal efficiency, REreponed, %
Emission stream flow rate, Qe, scfm
Temperature of emission stream, Te, °F
HAP
HAP concentration, HAPe, ppmv
Moisture content, Me, %
Temperature of condensation, Tcon. °F
Coolant used
Inlet temperature of coolant, TCOOi,i, °F
Coolant flow rate, QCOoiant, Ib/hr
Refrigeration capacity, Ref, tons
Condenser surface area, Acon,  ft2

4.8.2 Pretreatment of the Emission Stream
If water vapor is present in the emission stream, ice
may form on the condenser tubes when  coolants
such as chilled water or brine solutions are used,
decreasing the heat transfer efficiency and thus
lowering the  condenser's  removal efficiency. In
such cases,  dehumidifying  the emission stream is
necessary. This can be carried out in a  heat ex-
changer prior to the condenser.
                 Example Case
  Since the moisture content of the emission
  stream is negligible, no  pretreatment is neces-
  sary.
4.8.3 Condenser System Design Variables
The key design variable in condenser system de-
sign is the required condensation temperature for a
given removal efficiency or outlet concentration. A
condenser's removal efficiency depends on the na-
ture and concentration of emission stream compo-
nents. For example, compounds with high boiling
points (i.e.,  low volatility) condense more readily
compared to those with low  boiling  points. As-
sume, as a  conservative starting  point, that con-
densation will be considered as a HAP emission
control technique  for VOC's with boiling  points
above 100°F.

The temperature necessary to achieve a given re-
moval efficiency (or outlet concentration) depends
on the vapor pressure of the HAP in question at the
vapor/liquid equilibrium. Once the removal  effi-
ciency for a given HAP is specified, the required
temperature for condensation can be determined
from data on its vapor pressure-temperature rela-
tionship. Vapor pressure-temperature data can be
represented graphically (Cox charts) as shown  in
Figure  4-19for typical VOC's. The coolant selection
is then based on the condensation temperature re-
quired. See Table 4-10 for a summary of practical
limits for coolant selection.

In a permit evaluation, use Table 4-10 to determine
if the values reported for the condensation tem-
perature (Tcon) and the type of coolant selected are
consistent. Also, check if the coolant inlet tempera-
ture is based on a reasonable approach tempera-
ture (a conservative value of  15°F is used in the
Figure 4-19.  Vapor pressure-temperature relationship.
                          (1) Styrene
                          (2) Toluene
                          (3) Ethylene dichloride
                          (4) Hexane
  0.1
  0.00125  I   0.00175   I   0.00225   I   0.00275
        0.00150     0.00200     0.00250     0.00300

                [1/(Tcon + 460)] d/°R)
                      76

-------
Table 4-10.   Coolant Selection
   Required
 Condensation
  Temperature
   Icon (°F)a
Coolant
  Coolant
Temperature
Tcooi, <°F)b
  Tcon:c 60-80            Water           Tcon-15
 60>Tcon>45         Chilled water         Tcon-15

 45>Tcona-30      Brine solutions (e.g.,      Tcon-15
                  calcium chloride,
                   ethylene glycol)
                 Chlorofluorocarbons      Tcon-15
                   (e.g., Freon-12)
aAlso emission stream outlet temperature.
"Assume the approach as 15°F.
cSummer limit.
table). If the reported values are appropriate, pro-
ceed  with the calculations.  Otherwise,  the  appli-
cant's design is  considered unacceptable. The re-
viewer may then follow the calculation  procedure
outlined below.

4.8.4  Determining Condenser System Design
      Variables
The condenser system evaluated in this handbook
consists of a shell-and-tube heat exchanger with
the hot fluid (emission stream) in the shell side and
the cold fluid (coolant) in the tube side. The emis-
sion stream is assumed to consist of a two-compo-
nent  mixture: one condensible component  (HAP)
and one noncondensible component (air). Typical-
ly,  condensation for  such a system occurs non-
isothermally. To simplify the calculations, it is as-
sumed that condensation occurs isothermally.

4.8.4.1 Estimating Condensation Temperature
In the following calculations, it is assumed that the
emission stream entering the condenser  consists
of air saturated with the HAP in question. Calcula-
tions  for cases  involving  mixtures of  HAP's and
supersaturated streams are quite complex and will
not be treated  here; for additional information,
consult references 25 and 26.

For a  given removal efficiency, the first step in the
calculation procedure is to determine the concen-
tration at the outlet of the condenser. Use the fol-
lowing expression:

Ppartial  = 760 {(1 - 0.01  RE)/[(1 - (RE X 10~8
          x HAPe)]} HAPe x10~6           (4.8-1)

where Ppartjai is the partial pressure (mm Hg) of the
HAP in the exit stream  assuming the pressure in
the condenser is constant and at atmospheric. At
equilibrium between the gas and liquid phases, the
partial pressure of the HAP is equal  to its vapor
pressure at that temperature. Therefore, by deter-
mining the temperature at which this condition oc-
curs, the condensation temperature (Tcon) can be
specified. To carry out this calculation, vapor pres-
sure-temperature data for the  specific HAP are re-
quired (see Figure 4-19). Such data can be obtained
from references 18 and 27.
                                              Example Case
                              Using Equation 4.8-1 and Figure 4-19:
                              HAPe
                              RE
                              p
                              'partial
                       = 13,000 ppmv (styrene)
                       = 90%
                       = 760{[1-(0.01 x90)]/[(1-(90x1Q-8
                          x 13,000)]} 13,000 x10'6
                       = 1.0 mm Hg
                              'partial

                              For styrene, the value of [1/(Tcon  + 460)] corre-
                              sponding to 1.0 mm Hg in Figure 4-19 is about
                              0.00208.

                              Solving forTcon;
                             4.8.4.2 Selecting the Coolant
                             The next step is to select the coolant based on the
                             condensation temperature required. Use Table 4-
                             10 to specify the coolant type. For additional infor-
                             mation on coolants and their properties, see refer-
                             ences 18 and 27.
                                              Example Case
                              Based on Tcon = 20°F, the appropriate coolant is a
                              brine solution. Assume the brine solution is a
                              29% (wt) calcium chloride solution which can be
                              cooled down to -40°F (see reference 18).
                             4.8.4.3 Condenser Heat Load
                             Condenser heat load is defined as the quantity of
                             heat that  must be extracted  from the emission
                             stream to  achieve a certain  level of removal. It is
                             determined from  an energy balance, taking into
                             account the heat of condensation of the HAP, sensi-
                             ble heat change of the HAP, and the sensible heat
                             change in  the emission  stream. The calculation
                             steps are outlined below:

                             1.  a. Calculate moles of HAP in the inlet emission
                                  stream (Basis: 1 min):

                                  HAPe,m  = (Qe/387) HAPe x 10~6       (4.8-2)

                                  The factor 387 is the volume (ft3) occupied by
                                  1 Ib-mole of ideal gas at standard conditions
                                  (70°Fand 1 atm).
                                                                        77

-------
   b.  Calculate moles of HAP remaining in the out-
      let emission stream (Basis: 1 min):
HAPom   = (Qe/387)[1 - (HAPex
           [Pvapor/(Pe - Pvapor)]

where Pvapor is equal to Ppartiai
                                         (4.8-3)
   c. Calculate moles of HAP condensed (Basis: 1
     min):
     HAPcon = HAPe.m - HAP
                   e,m
                            o,m
                                    (4.8-4)
2.  a. Determine the  HAP's heat of vaporization
     (AH): Typically the heat of vaporization will
     vary  with  temperature. Using vapor pres-
     sure-temperature data as shown in Figure 4-
     19, AH can be estimated by linear regression
     for the  vapor  pressure and temperature
     range of interest. (See references  18 and  19
     for details.)

   b. Calculate  the  enthalpy change  associated
     with the condensed HAP (Basis: 1 min):
      = HAPcon [A H +
                                 (Te - Tcon)]
                                         (4.8-5)

      where CPHAP is the average specific heat of
      the HAP for the temperature interval Tcon - Te
      (Btu/lb-mole-°F).

   c.  Calculate the  enthalpy  change associated
      with the uncondensed HAP (Basis: 1 min):
                           (Te - Tc
                                    (4.8-6)
   d.  Calculate the  enthalpy change associated
      with  the noncondensible vapors  (i.e.,  air)
      (Basis: 1 min):
      Hnoncon   = [JQe/387) - HAPe,J
                 Cpair (Te - Tcon)
                                    (4.8-7)
      where Cpair is the average specific heat of air
      for the temperature interval Tcon - Te (Btu/lb-
      mole-°F).
3.  a.  Calculate the condenser heat load  (Btu/hr) by
      combining Equations 4.8-5, -6, and -7:
H
       ,oad
            = 1. 1x60
I  |_1      i U     \
'  nuncon ~ nnoncon'
                                         (4.8-8)
     The factor 1.1 is included as a safety factor.

4.8.4.4.  Condenser Size
Condenser systems are typically sized based on the
total heat load  and the overall heat transfer coeffi-
cient estimated from individual heat transfer + co-
efficients of the gas  stream and the coolant. An
accurate estimate  of individual coefficients can be
made using physical/chemical property data for the
gas stream, the coolant, and the specific shell-and-
tube system to be  used. Since this level of detail is
not appropriate here,  the value used for the overall
heat transfer coefficient is a conservative estimate.
                 Example Case
 Using Equations 4.8-2 to -8:
 1.  a.  Qe       =  2,000 scfm
       HAPe
       HAPP
                                                      b. P,
                                                           vapor
       Pe
       HAP0
       HAP

    c.  HAPr
                                                             o,m
                                       =  (2,000/387) 13,000 x 10  6
                                       =  0.06718 Ib-moles/min

                                       =  1.0 mm Hg
                                       =  760 mm Hg
                                       =  5.1008 [1.0/(760-1.0)]
                                       =  0.00672 Ib-moles/min

                                       =  .06718-0.00672
                                       =  0.0605 Ib-moles/min
                        2. a.  AH      =  17,445 Btu/lb-mole
                               (see Appendix B.10, reference 8, or refer-
                              ences 18 and 19)

                           b.  MWHAP  =  104.2 Ib/lb-mole
                              CPHAP    =  24 Btu/lb-mole-°F
                               (extrapolated from data in reference 27)
                              Hcon     =  0.0605(17,445 + 24(90-20)]
                              Hcon     =  1,157Btu/min

                           c.  Huncon    =  0.00672x24x(90-20)
                              Huncon    =  11.3Btu/min

                           d.  Cpair     =  6.96 Btu/lb-mole-°F
                               (see reference 3 for details)
                              Hn0ncon  =  ((2,000/387) - 0.06718] 6.96 x
                                           (90-20)
                              Hnoncon  =  2,485 Btu/min

                                       =  1.1 x 60 (1,157 + 11.3 +
                                           2,485)
                               H|0ad    =  241,100 Btu/hr
                                                    3.  a.  H
         |0ad
For additional information on how to calculate indi-
vidual  heat transfer coefficients, consult reference
25.
To size condensers, use the following equation to
determine the required heat transfer area:

Aeon = H,oad/UATLM                       (4.8-9)
where:
ACOn   =  condenser (heat exchanger) surface area,
         ft2
U      =  overall heat transfer coefficient, Btu/hr-ft2-
         °F
ATLM  =  logarithmic mean temperature  difference,
         °F
                                             and:

                                             ATLM =
             ' ' e    ' cool.o'    ' I con   ' i
                                                    con   ' cool.i'
                                                        In [(Te   TCOO| 0)   (Tcon   TCOO|])J
                                             where:
                                             Te      = emission stream temperature, °F
                                             TCOOI.O   = coolant outlet temperature, °F
                                             Tcon    = condensation temperature, °F
                                             TCooi,i   = coolant inlet temperature, °F
                      78

-------
 Assume that the approach temperature at the con-
 denser exit is 15°F. In other words, TCOO|, = (Tcon -
 15). Also, the temperature rise of the coolant fluid is
 specified as 25°F, i.e., TCOO|,0 = (TCOO|,0 + 25) where
 TCOOI.O is the coolant exit temperature. In estimating
 Acon, the overall  heat  transfer coefficient can  be
 conservatively assumed as 20 Btu/hr-ft2-°F; the ac-
 tual value will depend on the specific system under
 consideration. This is  based  on reference  26 in
 which guidelines on typical overall heat transfer
 coefficients for condensing vapor-liquid media are
 reported.

Example Case

Using Equation 4.8-9:
T6
' con
' cool, i
' cool.o
ATLM
ATLM
Hload
u
"con
"con
= SOT
= 20°F
= 20- 15 = 5°F
= TCOO|,i + 25 = SOT
f (90 - 30) - (20 -
[ In [(90 - 30)/(20 -
= 32°F
= 241,100Btu/hr
= 20 Btu/hr-ft2-°F
= 241,100/(20x32)
= 375 ft2




5)1
5)] J





4.8.4.5 Coolant Flow Rate
The quantity of  heat extracted from the emission
stream is transferred to the coolant. By a simple
energy balance, the flow rate of the coolant can be
calculated as follows:

^coolant = H|oacj/[Cpcoo|ant (TCOO|/0 - Tcoo|j)]   (4.8-10)

where:
           = coolant flow rate, Ib/hr
           = average specific heat of the  coolant
             over the temperature interval TCOO| ,
             to TCOO|,0, Btu/lb-°F

Specific  heat data for coolants are available  in re-
ferences 18 and 27.
C_P
   cooiant
                  Example Case
  Using Equation 4.8-10:
  Hload

  Tcool.i

  ' cool.o

  ^Pcoolant

  '-'coolant

  '-'coolant
            = 241,100 Btu/hr
            = 5°F
            = SOT
            = 0.65 Btu/lb-T (reference 18)
            = 241,100/[0.65(35-10)]
            = 14,840 Ib/hr
4.8.4.6 Refrigeration Capacity
A refrigeration unit is assumed to supply the cool-
ant at the required temperature to the condenser.
                                                    For costing purposes, the required refrigeration ca-
                                                    pacity is expressed in terms of refrigeration tons as
                                                    follows:

                                                    Ref= Hioad/12,000                      (4.8-11)
                                                    where Ref is the refrigeration capacity, tons.
                                                                     Example Case
                                                     Using Equation 4.8-11:
                                                     H|0ad  = 241,100 Btu/hr
                                                     Ref   = 241,100/12,000
                                                     Ref   = 20 tons
                                                    4.8.4.7 Recovered Product
                                                    To calculate costs, the quantity of recovered prod-
                                                    uct that can be sold and/or recycled to the process
                                                    must be determined. Use the following equation:

                                                    Qrec = 60 x HAPcon x MWHAp             (4.8-12)

                                                    where Qrec is the quantity  of product recovered,
                                                    Ib/hr.
                                                                     Example Case
                                                     Using Equation 4.8-12:
                                                     HAPcon   = 0.0605 Ib-moles/min
                                                     MWHAp  = 104.2lb/lb-mole
                                                     Qrec     =60x0.06065x104.2
                                                     Qrec     = 378 Ib/hr
4.8.5 Evaluation of Permit Application
Compare the results from the calculations and the
values supplied by the permit applicant using Table
4-11. The calculated values in the table are based
on the Example Case.  If the calculated  values  of
Tcon, coolant type, Acon, QCOO|ant, Ref, and Qrec are
different from the reported  values for these varia-
bles, the differences may be  due to the assump-
tions involved in the calculations. Therefore, the
reviewer may wish to  discuss the  details  of the
proposed design with the permit applicant.
Table 4-11.   Comparison of Calculated Values and Values
            Supplied by the Permit Applicant for
            Condensation

Condensation temperature, Tcon
Coolant type
Coolant flow rate, Qcoolam
Condenser surface area, Acon
Refrigeration capacity, Ref
Recovered product, Qrec
Calculated
Value
(Example Case)8
20°F
Brine solution
14,840 Ib/hr
375ft2
20 tons
378 Ib/hr
Reported
Value

                                                   3 Based on Emission Stream 6.
                                                                          79

-------
If the calculated values agree with the reported
values, then the design and operation of the pro-
posed condenser system may be considered ap-
propriate based on the assumptions made in this
handbook.
4.9  Fabric Filters
Fabric filter collectors (also known as baghouses)
are one of the most efficient means of separating
particulate matter from a gas stream. Fabric filters
are capable of maintaining mass collection efficien-
cies  of greater than 99 percent down to a particle
size  approaching 0.3  jjim in  most applica-
tions.(28,29,30) This efficiency is largely insensitive
to the physical characteristics of the gas and dust,
and, depending on fabric cleaning method, to the
inlet dust loading.(28,31) Physical limitations of the
fabric materials to the temperature, moisture con-
tent, and corrosivity of the gas  stream reduce the
applicability of fabric filters. Variables  considered
in baghouse  design include fabric type, cleaning
method, air-to-cloth ratio, and equipment configu-
ration. The filter fabric, cleaning method, and air-
to-cloth ratio all should be selected concurrently;
choice  of these parameters  is mutually  depen-
dent.(28) Equipment configuration is of secondary
importance unless  site-specific space limitations
exist that require configuration to be  of primary
importance in the fabric filter design.

Fabric filter systems typically are  designed on the
basis of empirical information  obtained through
testing and long-term  actual operating  experience
for similar combinations of cleaning method, fabric
type, and dust  rather than by analytical  meth-
ods.(28) Although theoretical  equations exist to
predict the performance of filtering systems under
various conditions, these equations are not very
useful as design tools.  Therefore,  discussion of
baghouse design in this  section provides qualita-
tive  guidance rather  than predictive  equations.
Generally, fabric filter design for HAP's is no differ-
ent than fabric filter design for control of any other
type of particulate matter. However, due to the haz-
ards associated with HAP's, greater care must be
taken to ensure that control is consistently of high
efficiency and that the control device is leak-proof,
thus  preventing accidental  release of the gas
stream and captured pollutants. For these reasons,
design of a fabric filter for HAP's should only con-
sider selected fabric cleaning methods, and the de-
sign  should  specify an induced  draft  fan (i.e., a
negative pressure or suction baghouse) rather than
a forced draft fan (i.e., a positive pressure  bagh-
ouse). Information presented in this section can be
used to provide guidance for or to evaluate the
appropriateness of baghouse design  for certain
HAP applications.
Appendix C.9 provides a worksheet to record the
information obtained during the performance  of
the fabric filter design procedures.

4.9.1 Data Required
The data necessary to perform the  design steps
consist of the HAP emission stream  characteristic
previously compiled on the  HAP Emission Stream
Data Form and the required HAP control as deter-
mined by the applicable regulations.
                 Example Case
  Fabric filtration was one of the selected control
  techniques  for the  municipal  incinerator. The
  pertinent data for these procedures are found on
  the HAP Emission Stream Data Form (see Figure
  3-8).

  Flow rate, Qe/a = 110,000 acfm
  Moisture content, Me = 5% vol
  Temperature, Te =  400°F
  Particle mean diam. = 1.0 (Jtm
  S03 content = 200  ppm (vol)
  Particulate content  = 3.2 grains/scf - flyash
  HAP content = 10% (mass) cadmium
In the case of a permit review for a fabric filter, the
following data should be supplied by the applicant:
Filter fabric material
Cleaning method
Air-to-cloth ratio, ft/min
Baghouse configuration

The design criteria and considerations discussed in
this section will be used to evaluate the reasonable-
ness of the applicant's proposed design.

4.9.2 Pretreatment of the Emission Stream
As  discussed  in Section  3.3.1, the temperature of
the emission stream should be within 50° to 100°F
above the stream dew point. Procedures for deter-
mining the dew point of an emission  stream are
provided in Appendix B.1.  If the emission stream
temperature does not fall within the stated range,
pretreatment (i.e., emission stream preheating or
cooling)  is necessary. Methods of pretreatment are
briefly discussed in reference 32. If pretreatment is
performed, the emission stream characteristics will
be  altered. The primary characteristics affecting
baghouse design are emission stream temperature
and flow rate. Therefore, after selecting a tempera-
ture for the emission stream, the new stream flow
rate must  be calculated. The calculation  method
depends upon the type of pretreatment performed;
use appropriate standard industrial equations. The
use of pretreatment mechanical  dust collectors
may also be appropriate.  If the emission  stream
contains an appreciable  amount of large particles
                      80

-------
(20 to 30 |xm), pretreatment with mechanical dust
collectors is typically performed. (Appendix B.11,
reference 8, further describes the use of mechani-
cal dust collectors.)

4.9.3 Fabric Filter System Design Variables
Successful design of a fabric filter depends on the
proper selection of fabric and cleaning method and
on an  adequate air-to-cloth ratio. All fabric  filter
systems share the same basic features and operate
using the principle of aerodynamic capture of parti-
cles by fibers. Systems vary,  however, in certain
key details of construction  and in the operating
parameters. The design variables of particular in-
terest are filter bag material, fabric cleaning meth-
od, air-to-cloth ratio, baghouse configuration, and
materials of construction.

As stated earlier, the first three variables should be
considered  concurrently. The  configuration and
construction  materials are important, but secon-
dary, considerations.  The following  subsections
discuss step-by-step procedures for selecting  each
of these design  variables as they may apply to a
specific particulate HAP control situation. Because
HAP control is similar to particulate control in gen-
eral, a good verification of these procedures can be
accomplished by consulting the section about the
particular industry in a document entitled Control
Techniques for Particulate Emissions from  Station-
ary Sources—Volume 2, or in the Mcllvaine Fabric
Filter Manual.(31,33) (Note:  Because these design
variables are considered concurrently, the Example
Case is presented at the end of Section 4.9.3.)


4.9.3.1 Fabric Type
Several types of natural and  synthetic fabric are
used in baghouse systems. Gas stream characteris-
tics such as temperature, acidity,  alkalinity, and
particulate matter properties (e.g., abrasiveness
and hygroscoposity), determine the fabric type to
be used.(28,34)  In  many instances, several fabric
types will be appropriate, and a final selection will
be chosen only when cleaning method  and the
desired air-to-cloth ratio are considered.
Most of the  principal  synthetic fibers have been
adapted for use as filtering  fabrics  while the only
natural fibers in common use are cotton and wool.
Some of the more common synthetic fibers in com-
mercial use are nylon (aromatic and polyamide),
acrylic,  polyester, polypropylene,  fiberglass, and
fluorocarbon.  Natural  fibers can be used for gas
temperatures up to 200°F and have  only moderate
resistance to acids and alkalis contained in the gas
stream.(28,31,34) Synthetics can operate  at  tem-
peratures up to  550°F and generally have greater
chemical resistance.(28,31,34) Therefore, while the
initial cost of the  synthetic  filter fabric is greater
than the cost for natural fibers, the increased ser-
vice life and improved operating characteristics of
the synthetics make them a preferred choice in a
wide range of industrial situations.

Almost all of the filter fabrics can be constructed in
either  a woven or a felted manner (cotton and fi-
berglass can be constructed in a  woven manner
only). Woven fabrics are made up of yarn in one of
a variety of patterns that allow spaces between the
fibers, whereas felted fabrics are  composed of a
thick mat of randomly oriented fibers. When woven
fabrics are new, particles penetrate the pores of the
fabric fairly easily. As filtering continues, however,
more particles are retained on the filter threads and
on the particles already collected. As this dust layer
or "cake" builds up, particle penetration drops to a
very low level. Cleaning of woven  fabrics must be
performed so that a layer of this dust cake remains
on the fabric, enabling particle penetration to re-
main low.(28,33) Felted fabrics are thick enough
that  a  dust cake does not need to remain  on the
fabric  in order to maintain a good collection effi-
ciency.(28,31) This difference between woven and
felted fabrics has important implications for selec-
tion of fabric cleaning method, as described in Sec-
tion 4.9.3.2. Felted fabrics are more expensive than
woven fabrics.

Table 4-12 presents information on the maximum
continuous operating temperature and resistance
characteristics of  commonly  used filter fabrics.
Knowing the emission stream  characteristics, Ta-
ble 4-12 can be used to select an appropriate fabric
filter type (or types). Although the information pre-
sented is qualitative,  Table 4-12 provides a good
basis either for selecting a fabric or for evaluating
the appropriateness of a fabric in a permit applica-
tion.

When a number of fabrics are suitable for an appli-
cation, the relative  cost of the fabrics may be the
key decision criterion. In general, fluorocarbon and
nylon aromatic bags are the most expensive, fol-
lowed  by wool and fiberglass. The  remaining com-
monly used synthetics are generally less expensive
than fiberglass (polypropylene, polyester, acrylic,
nylon polyamide, and modacrylic), while cotton is
generally the least expensive fabric.(29,30,31,35)

4.9.3.2 Cleaning Method
As dust accumulates on the filtering elements, the
pressure drop  across the bag compartment in-
creases until cleaning of the bags  occurs. A timer
can be used to control the cleaning cycle or pres-
sure drop can be monitored so that cleaning occurs
when some maximum desirable value is reached.
At this point the  bags in the compartment  are
cleaned to remove the collected dust and the cycle
is then repeated. The two basic mechanisms used
to accomplish bag cleaning are flexing of the fabric
to break up and dislodge the  dust cake, and re-
                                                                        81

-------
Table 4-12.   Characteristics of Several Fibers Used in Fabric Filtration8
                                                                  Resistance0
Fiber
Type"
Cottond
Wool8
Modacrylic8
(Dynel)
Polypropylene6
Nylon Polyamide6
(Nylon 6 & 66)
Acrylic6
(Orion)
Polyester6
(Dacron, Creslan)
Nylon Aromatic6
(Nomex)
Fluorocarbon6
(Teflon, TFE)
Fiberglassd
Max. Continuous
Operating
Temp. (°F)
180
200
175

200
220

260

275

450

500

550
Abrasion
G
F/G
F

E
e

G

E

E

F/G

P,Gh
Mineral
Acids
P
F
VG

E
P

G

G

F

E9

VG
Organic
Acids
G
F
VG

E
F

G

G

G

E9

E
Alkalies
G
P/F
G

E
VG

F

G

VG

E9

P
Solvent
E
G
G

G
E

E

E

E

E9

E
"References 8, 29, 31, 34, 35, and 37. Where data differed, a representative category was chosen.
bRepresents the major categories of filtration fibers. Names in parentheses indicate some principal trade names.
CP = poor resistance, F = fair resistance, G = good resistance, VG = very good resistance, and E = excellent resistance.
dWoven fabrics only.
"Woven or felted fabrics.
'Considered to surpass all other fibers in abrasion resistance.
9The most chemically resistant of all these fibers.
hAfter treatment with a lubricant coating.
versed air flow through the fabric to remove the
dust.(28) These may be used separately or in con-
junction with  one another. The three principal
methods  used to accomplish fabric cleaning are
mechanical shaking (manual or automatic), reverse
air flow, and  pulse-jet cleaning. The first method
uses only the fabric flexing mechanism; the latter
two methods  use a combination of the reverse air
flow and fabric flexing mechanisms.

Selection of a cleaning method is based on the type
of fabric used, the pollutant collected, and the man-
ufacturer's, vendor's, and industry's experiences. A
poor  combination of filter fabric  and cleaning
method can cause premature failure of the fabric,
incomplete cleaning, or  blinding of the fabric.(28)
Blinding of a  filter fabric occurs when the fabric
pores are  blocked and effective cleaning can not
occur. Blinding can result because moisture blocks
the pores or increases the adhesion of the dust, or
because a high  velocity gas stream imbeds the
particles too deeply in the fabric.(28) The selection
of a cleaning method may be based on cost, espe-
cially where more than one method is applicable.
Table  4-13 contains  a  comparison of cleaning
methods. Cleaning methods are discussed individ-
ually below.

With mechanical shaking,  bags are hung  on an
oscillating framework that periodically shakes the
bags at timed intervals or at a predefined pressure
drop level.(28,30,34) The shaker mechanisms pro-
duce a violent action on the fabric filter bags and, in
general, produce more fabric wear than the other
types of cleaning mechanisms.(30) For this reason,
mechanical shaking is used  in conjunction with
heavier and more durable fabric materials, such as
most woven fibers.(30,38) Bags with  poor or  fair
abrasion ratings  in Table 4-12 (such as fiberglass)
should not be chosen for fabric filters cleaned by
mechanical shaking unless they are treated with a
special coating  before use.  Although shaking is
abrasive to the fabric, it does allow a dust cake to
remain on the fabric, thus maintaining a high  col-
lection efficiency.

Bags are usually  taken off-line for cleaning by me-
chanical shaking  so that no gas flows through  the
bags being cleaned. Thus, reentrainment of parti-
cles is minimized.  Because dust dislodgement is
not severe (i.e., a light dust cake remains on  the
fabric), and because cleaning occurs off-line, outlet
concentrations are almost constant with varying
inlet dust loading and through entire cleaning cy-
cles when using mechanical  shaking.(28) Further
control efficiency is very high, and, in fact, properly
selected woven  fabrics cleaned by  mechanical
shaking can  provide much greater particle collec-
tion than  pulse-jet cleaned felted fabrics in  many
applications.(31)  For these  reasons,  mechanical
shaking is a good method to clean fabric filters
controlling emissions containing HAP's.(31)

Reverse air flow cleaning is used to flex or collapse
the filter bags by allowing a  large volume of  low
pressure air to pass countercurrent to the direction
                       82

-------
 Table 4-13.    Comparisons of Fabric Filter Bag Cleaning Methods (31)
                                                         Cleaning Method
Parameter
Cleaning On- or Off-line
Cleaning Time
Cleaning Uniformity
Bag Attrition
Equipment Ruggedness
Fabric Type
Filter Velocity
Power Cost
Dust Loading
Maximum Temperature3
Collection Efficiency
Mechanical
Shake
Off-line
High
Average
Average
Average
Woven
Average
Low
Average
High
Good
Reverse
Airflow
Off-line
High
Good
Low
Good
Woven
Average
Low to medium
Average
High
Good
Pulse-jet
Individual
Bags
On-line
Low
Average
Average
Good
Felt
High
High
Very high
Medium
Lower
Pulse-jet
Compartmented
Bags
Off-line
Low
Good
Low
Good
Felt
High
Medium
High
Medium
Lower
aFabric limited.
of normal gas stream flow during filtration.(30,34)
Reverse air is provided either by a separate fan or
by a vent in the fan damper, which allows a back-
wash of air to clean the fabric filters.(30,34) Reverse
air flow cleaning usually occurs off-line. Reverse air
cleaning allows the use of fragile bags, such as
fiberglass, or light-weight bags, and usually results
in longer  life for the bags.(30) As with mechanical
shaking, woven  fabrics are  used,  and  because
cleaning is less violent than with pulse-jet cleaning
and  occurs off-line, outlet concentrations are al-
most constant with varying inlet dust loading and
throughout the cleaning cycle.  Reverse  air flow
cleaning  is, therefore, a good choice for fabric
cleaning in HAP control situations.

In pulse-jet cleaning, a high pressure air pulse is
introduced  into the bag from the top through a
compressed air jet.(30,34) This rapidly expands the
bag,  dislodging the particles.  Thus, the fabric is
cleaned thoroughly through a vibration effect. The
pulse of air cleans  so effectively that no dust cake
remains on the fabric to contribute to  particulate
collection. Because such a cake is essential for ef-
fective  collection on woven  fabrics, felted fabrics
are generally used in pulse-jet cleaned fabric fil-
ters.(28) All of the fabric materials may  be used
with pulse-jet cleaning except cotton or fiberglass.

Because the cleaning air pulse is of such high pres-
sure (up to 100 psi) and short duration (<0.1 sec),
cleaning is  usually accomplished on-line. Extra
bags are not necessary, therefore, to compensate
for bags off-line during cleaning. Cleaning occurs
more frequently than with  mechanical shaking or
reverse air flow cleaning, which permits higher air
velocities  (higher A/C ratios)  than the other clean-
ing methods. Further, because the bags move less
during cleaning, they may be packed more closely
together.  In combination, these features allow
pulse-jet cleaned fabric filters to be installed in a
smaller space, and thus, at a lower cost, than fabric
filters cleaned  by the  other methods.(28,34) This
cost savings may be somewhat counterbalanced
by the greater expense and more frequent replace-
ment required of felted bags, the higher power use
that may occur, and the  installation of the fabric
filter framework  that  pulse-jet cleaning re-
quires.(28,34)

Pulse-jet cleaning is not,  however, recommended
for HAP control situations for several reasons. First,
because cleaning occurs  on-line, the  rapid  high-
pressure pulse generated during cleaning causes
increased emissions from the bags. Mass emis-
sions can vary  by as  much as 100 times over a
filtration cycle.(31) Second, although collection effi-
ciencies of pulse-jet cleaned fabric filters are in the
99.9 to 99.99 percent range, filtering efficiency of
pulse-jet cleaned filters is inferior to that of me-
chanically shaken or reverse air cleaned filters that
have a good cake buildup.(31) In one study, aver-
age outlet concentrations were two to three orders
of magnitude higher for  pulse-jet cleaned filters
than  for  mechanically shaken filters.(31)  Third,
emissions from pulse-jet systems are strongly de-
pendent on the inlet concentration; thus, the collec-
tion efficiency  rather than the effluent concentra-
tion tends to be relatively  constant for fabric filters
using pulse-jet cleaning.(28,31) For these  reasons,
outlet emission levels are  not as constant or as low
when using pulse-jet cleaning as when using either
mechanical  shaking or reverse air flow  cleaning.
Pulse-jet cleaning is, therefore, not recommended
for fabric filters used in HAP control situations or
for high inlet loadings involving fine particulate
matter.(29)

In cases of permit evaluation where pulse-jet clean-
ing is believed to be  adequate  to meet specific
regulations in specific applications, several options
are available  to minimize the disadvantages of
                                                                         83

-------
pulse-jet cleaning. First, pulse-jet filter bags can be
compartmentalized to permit off-line cleaning; ad-
ditional bags must be installed to allow this.(31)
Second, reduced filtration velocity, or pulse inten-
sity, will decrease average outlet concentration.(31)
Third, bags should be flexible, lightweight, and in-
elastic, with uniform pore structure, to obtain maxi-
mum particle collection.(31) These changes, in ef-
fect, alter the typical pulse-jet baghouse such that it
behaves (i.e., cleans the bags) in a manner that is
very similar to that of a reverse air baghouse.
4.9.3.3 Air-to-Cloth Ratio
The air-to-cloth (A/C) ratio, or filtration velocity, is a
traditional fabric filter design parameter defined as
the actual volumetric flow rate (acfm) divided by
the total active, or net, fabric area (ft2). The  A/C
ratio is an important indicator of the amount of air
that can be filtered in a given time when consider-
ing the dust to be collected, cleaning method  and
fabric to be used, and the characteristics of the gas
stream to be filtered for an individual situation.
Selection of an appropriate range of A/C ratios is
not based on any theoretical or empirical relation-
ship, but rather is based on industry and fabric filter
vendor experience from actual fabric filter installa-
tions. A ratio is usually recommended for a specific
dust and a  specific  cleaning method. For  typical
design calculations, the A/C ratio must be obtained
from the literature or the manufacturer.

Table 4-14 summarizes the ranges of recommend-
ed A/C ratios by typical bag cleaning method for
many dusts and fumes. These ranges are meant to
serve as a guide; A/C ratios may vary from those
reported. Fabric filter size and cost will vary with
A/C ratio; lower A/C ratios, for example, will require
that a larger and  more expensive fabric filter be
installed. (Note: Pulse-jet cleaning is  not recom-
mended for HAP control situations;  the A/C ratio
for control of streams containing HAP's will, there-
fore, be fairly low.) In addition to evaluating a par-
ticular fabric filter application, the A/C ratio and the
emission stream flow rate (Qe,a) are used to calcu-
late net cloth area (Anc):
  Qe
A/C ratio

where:
Qe,a
          =  An
                                         (4.9-1)
         =  emission stream flow  rate at actual
            conditions acfm
A/C ratio =  air-to-cloth ratio, acfm/ft2 or ft/min
Anc      =  net cloth area, ft2
Net cloth area is the cloth area in active use at any
point in time. Gross cloth area (Atc), by compari-
son, is the  total cloth area  contained  in  a fabric
                                                  filter, including that which is out of service at any
                                                  point in time for cleaning or maintenance. In this
                                                  handbook, costing of the fabric filter structure uses
                                                  net cloth area, while costing of fabric filter bags
                                                  uses gross cloth area. Table 4-15 presents factors
                                                  to obtain gross cloth area from net cloth area:
                                                   Anc x Factor = A,,
                                         (4.9-2)
where:
Factor = value from Table 4-15, dimensionless
Atc    = gross cloth area, ft2

Fabric filters with a higher A/C ratio require fewer
bags to  accomplish cleaning, and, therefore, re-
quire less space and may be less expensive. Other
costs, such as more expensive (felted) bags, bag
framework  structure,  use of increased pressure
drop and corresponding increased  power require-
ments, etc., may counterbalance to some degree
the savings of high A/C ratio systems.


4.9.3.4 Baghouse Configuration
The  basic configuration  of a baghouse varies ac-
cording to  whether the gases  are pushed through
the system by a fan located on the upstream  side
(forced draft fan), or pulled through by locating the
fan on the  downstream side (induced draft fan). A
baghouse using forced draft fans is called a posi-
tive-pressure baghouse; one  using induced draft
fans is called a negative-pressure or suction bag-
house. Positive-pressure baghouses may be either
open to the atmosphere or closed (sealed and pres-
sure-isolated from the atmosphere). Negative-pres-
sure  baghouses can only be  of the closed  type.
Only the closed suction design should be selected
for a HAP application to prevent accidental release
of captured pollutants.(34) The higher the gas
stream dew point, the greater the precaution that
must be  taken to prevent condensation, which can
moisten  the filter cake, plug the cloth, and promote
corrosion of the housing  and hoppers. In a suction-
type fabric filter, infiltration of ambient air can oc-
cur, which can lower the temperature below design
levels. Therefore, the structure walls and hoppers
of this type of baghouse should be insulated to
minimize the possibility of condensation.

4.9.3.5 Materials of Construction
The  most common  material  used  in fabric filter
construction is carbon steel. In cases where the gas
stream contains  high  concentrations of S03 or
where liquid-gas contact areas are involved, stain-
less  steel may be required.  Stainless steel will in-
crease the cost of the fabric filter significantly when
compared to carbon steel.(30) However, by keeping
the emission stream temperature above the dew
point and by insulating the baghouse, the use of
stainless steel should not be necessary.
                      84

-------
Table 4-14.   Recommended Air-to-Cloth (A/C) Ratios for Various Ousts and Fumes by Cleaning Method (28, 36)

                                       A/C Ratios Recommended for Cleaning Method (ft/min)
Dust
or Fume
Abrasives
Alumina
Aluminum
Aluminum Oxide
Asbestos
Bauxite
Blast Cleaning
Carbon
Carbon Black
Chrome
Coal
Coke
Dyes
Fertilizer
Flint
Fly Ash
Foundry
Glass
Graphite
Gypsum
Iron Ore
Iron Oxide
Iron Sulfate
Lead Oxide
Leather
Lime
Limestone
Machining
Manganese
Metal Fumes
Metal Powders
Mica
Paint Pigments
Paper
Perchlorates
Plastics
Polyethylene
PVC
Resin
Silica
Silica Flour
Silicates
Silicon Carbide
Slate
Starch
Talc
Shaker
2.0-3.0
2.25-3.0
3.0
2.0
2.5-4.0
2.25-3.2
3.0 - 3.5
1.2-2.5
1.5-2.5
1.5-2.5
2.0 - 3.0
2.5
2.0
2.0-3.5
2.5
2.0
#
2.5
1.5-3.0
2.0-3.5
2.0-3.5
2.0-3.0
2.0-2.5
2.0-2.5
3.5-4.0
2.0-3.0
2.0 - 3.3
3.0
2.25
1.5
2.0
2.25 - 3.3
2.0
3.5 - 4.0
*
2.0-3.0
#
#
2.0
2.25-2.8
2.0-2.5
*
*
2.5 - 4.0
2.25
2.25
Reverse Air
*
*
*
*
*
#
*
#
1.1 -1.5
#
#
*
*
1.8-2.0
*
2.1 -2.3
#
*
1.5-2.0
1.8-2.0
#
1.5-2.0
1.5-2.0
1.5-1.8
*
1.5-2.0
#
X
*
1.5-1.8
#
1.8-2.0
*
#
*
#
#
#
*
1.2-1.5
*
*
*
*
#
*
Pulse-Jet
9
#
16
*
9-16
8-10
*
5-7
8-12
9-12
12-16
9-12
10
8-10
#
9-10
8-12
*
7-9
10-16
11 -12
8-16
6-8
6-9
15-20
10-16
8-12
16
#
6-9
9-10
9-11
#
10-12
10
7-10
10
7
8-10
7-12
#
9-10
10
12- 14
*
#
*No information available.
4.9.4 Evaluation of Permit Application
Using  Table  4-16, compare the results from this
section and the data supplied by the permit appli-
cant. The calculated values are based on the exam-
ple case. As pointed out in the discussion on fabric
filter design  considerations,  the basic design pa-
rameters  are generally  selected without  the  in-
volved, analytical approach that characterizes
many other control systems, such as an absorber
system (Section 4.7). Therefore, in  evaluating the
reasonableness of any system specifications on a
permit application, the reviewer's main task will be
to examine each parameter in terms of its compati-
bility  with the gas  stream and  paniculate condi-
tions  and with the other selected parameters. The
following questions should be asked:

1.  Is the temperature of the emission stream enter-
   ing the baghouse within 50° to 100°F above the
   stream dew point?
2.  Is the selected fabric material compatible with
   the conditions of the emission stream; that is,
   temperature and composition (see Table 4-12)?
                                                                        85

-------
Table 4-15.    Factors to Obtain Gross Cloth Area from Net
            Cloth Area (30)
Net Cloth Area, Ano
(ft2)
1 - 4,000
4,001 - 12,000
12,001 - 24,000
24,001 - 36,000
36,001 - 48,000
48,001 - 60,000
60,001 - 72,000
72,001 - 84,000
84,001 - 96,000
96,001 - 108,000
108,001 - 132,000
132,001 - 180,000
180,001 +
Factorto Obtain
Gross Cloth Area, Atc
(ft2)
Multiply by 2
Multiply by 1.5
Multiply by 1.25
Multiply by 1.17
Multiply by 1.125
Multiply by 1.11
Multiply by 1.10
Multiply by 1.09
Multiply by 1.08
Multiply by 1.07
Multiply by 1.06
Multiply by 1.05
Multiply by 1.04
                 Example Case
 Table 4-12 indicates that filter fabrics that can
 withstand the 400°F emission stream tempera-
 ture are nylon  aromatic (Nomex), fluorocarbon
 (Teflon), and fiberglass. Because there is a high
 potential for acid damage (i.e., a high S03 con-
 tent), however, Nomex bags should not be con-
 sidered. Because  HAP's are present, only  me-
 chanical shaking  or reverse air flow cleaning
 methods are advisable.

 Using Table 4-14 for fly ash type dust, a low A/C
 ratio is expected for the two acceptable cleaning
 methods (2 to  2.3 ft/min). Because a fiberglass
 bag would provide the most protection during
 temperature surges,  and  because fiberglass
 bags may be less expensive, it may be the fabric
 of choice for an installation with these emissions
 characteristics. Fiberglass bags would require
 that reverse air cleaning be used. Teflon bags
 with mechanical shaking could  also be a  possi-
 bility. The documents that  describe experience
 in  certain industry applications  support the
 choice of fiberglass bags with  reverse air flow
 cleaning.(31,33)
Table 4-16.   Comparison of Calculated Values and Values
            Supplied by the Permit Applicant for Fabric
            Filters

                          Calculated Value   Reported
                          (Example Case)3    Value
Emission Stream Temp.
  Range"
Selected Fabric
  Material
Baghouse Cleaning
  Method

A/C ratio = Q"'a
          Anc
 Baghouse Configuration
365°-415°F
Fiberglass or
  Teflon
Mechanical shaking
  or reverse air flow

2-2.3 ft/min

Negative pressure
                           3.  Is the baghouse  cleaning  method compatible
                              with  the  selected fabric material  and its con-
                              struction; that is, material  type and woven or
                              felted construction (see Section 4.9.3.2 and Ta-
                              ble 4-13)?
                           4.  Will the selected  cleaning  mechanism provide
                              the desired control?
                           5.  Is the A/C ratio appropriate for the application;
                              that is, type of dust and cleaning method used
                              (see Table 4-14)?
                           6.  Are the values provided for the gas flow rate,
                              A/C ratio, and  net cloth area  consistent? The
                              values can  be checked with the following equa-
                              tion:
                              A/C ratio = §§•*-
                                         ^"*nc

                              where:
                              A/C ratio = air-to-cloth ratio, ft/min
                                                                                             (4.9-3)
 "Based on the municipal incinerator emission stream.
 bSee Section 3.3.1.
   Qe,a   = emission stream  flow rate  at  actual
           conditions, acfm
   Anc   = net cloth area, ft2

7. Is the baghouse configuration appropriate; that
   is, is it a negative-pressure baghouse?

A particular  manufacturer/customer combination
may employ somewhat different criteria in their
selection of design parameters (such as lower an-
nualized costs of operation at the expense of higher
initial costs), and so a departure from the "rules-of-
thumb"  discussed here may still be compatible
with achieving the needed high collection efficien-
cies. Further discussions with the permit applicant
are recommended to evaluate the design assump-
tions and to reconcile any apparent discrepancies
with usual practice.

4.3.5  Determination of Baghouse Operating
      Parameters
Many times, optimization of a fabric filter's collec-
tion efficiency occurs in the field after construction.
The following  discussion does not pertain to the
preliminary design of a fabric filtration control sys-
tem;  however, the information presented should
be helpful in achieving and maintaining the desired
collection efficiency for the installed  control sys-
tem.

4.9.5.1 Collection Efficiency
A well designed fabric filter can achieve collection
efficiencies in excess of 99 percent, although opti-
mal performance  of a fabric filter system may not
occur for a number of cleaning cycles as the new
filter material is "broken in." The fabric filter collec-
tion efficiency is related to the pressure drop across
the system, component life, filter fabric, cleaning
method and frequency, and A/C ratio. These factors
should be reevaluated if fabric filter performance is
less than  permitted. Modifications to improve per-
                       86

-------
formance include changing the A/C ratio, using a
different fabric, or replacing worn or leaking filter
bags. Collection efficiency can be improved by de-
creasing the frequency of cleaning or allowing the
system  to  operate  over a greater pressure drop
before cleaning is initiated.

4.9.5.2 System Pressure Drop
The pressure drop across the operating fabric filter
system is a function of the difficulty with which the
gas stream passes through the filter bags and accu-
mulating dust cake, how heavy the dust deposit is
prior to bag cleaning, how efficient cleaning is, and
if the  filter bags are plugged or blinded. Normally,
the value of this parameter is set at about 3  to 4
inches of water, although pressure drops in excess
of 10 inches have been used.(29) In actual oper-
ation, variations in pressure drop outside of the
design range may be indicative of problems within
the fabric filter system. Higher than expected pres-
sure differentials  may indicate: (1) an  increase in
gas stream volume; (2) blinding of the filter fabric;
(3) hoppers full of  dust, thus blocking the bags;
and/or (4) inoperative cleaning mechanism. Lower
than expected pressure  differentials may indicate:
(1) fan or motor problems, (2) broken or undamped
bags, (3) plugged inlet ducting or closed damper,
and/or (4)  leakage  between  sections  of the bag-
house.

As the dust cake  builds up during filtration, both
the collection efficiency  and system  pressure drop
increase. As the pressure drop increases toward a
maximum, the filter bags (or at least a group of the
bags contained in one isolated compartment) must
be cleaned to reduce the dust cake resistance.  This
cleaning must be timed and performed so as to
accomplish the following: (1) to keep the pressure
drop and thus operating costs, within  reasonable
limits; (2) to clean bags as gently and/or infrequent-
ly as possible to minimize bag wear and to maxi-
mize efficiency; and (3) to leave a sufficient dust
layer on the bags to maintain filter efficiency and to
keep the instantaneous A/C ratio immediately after
cleaning from reaching excessive levels.

In  practice, these various considerations are bal-
anced using engineering judgment  and field  trial
experience to optimize the total system operation.
Changes in process or in fabric condition through
fabric aging will cause  a shift in the cleaning re-
quirements of the system. This shift may require
more frequent manual adjustments to the automat-
ic  control to achieve the minimum cleaning re-
quirements.

4.10  Electrostatic Precipitators
Electrostatic precipitators (ESP's) use  an  electro-
static field to charge particulate matter contained in
the gas stream. The charged particles then migrate
to a grounded collecting surface. The collected par-
ticles are dislodged from the collector surface peri-
odically by vibrating or rapping the collector sur-
face, and subsequently collected in a hopper at the
bottom of the ESP.

There are two basic types of ESP's: single stage
and two stage.(32,37) In the single stage precipita-
tor, which may be wet or dry, ionization and collec-
tion are combined, whereas in the two stage preci-
pitator, ionization and collection  are  done in
separate steps. Dry, single stage ESP's are the most
common.  Wet electrostatic precipitators, while not
as common as dry ESP's, can be used to remove
both solid and gaseous pollutants.

The most  important variable considered in the de-
sign of an  ESP is collection plate area; this assumes
that the ESP  is provided with an optimum level of
secondary voltage.  Collection plate area is a func-
tion of the desired collection efficiency, gas stream
flow  rate  and particle  drift  velocity.(32,37,39,40)
Other design details to be estimated by the vendor
include (but are not limited to) expected secondary
voltage and current, electrical sections alignment,
and  direction of gas flow. In this document, an
approximate  method to size an ESP is given.

Particle drift  velocity is a complicated  function of
particle size, gas velocity, gas temperature, particle
resistivity, particle agglomerization, and the phys-
ical and chemical properties of the  particulate mat-
ter. The theoretical relationship of the drift velocity
to the variables is discussed extensively in the lit-
erature.(32,37,39,40) Unfortunately, there are no
empirical  equations readily available to  calculate
drift velocity  directly from  these variables. There-
fore, in determining drift velocity for a given emis-
sion stream,  equipment vendors often rely upon
historical  data for similar streams and  data estab-
lished from pilot plant tests. Published information
on drift velocity (based on design data for actual
installations to represent typical gas characteris-
tics) are available for several industrial emission
streams.(37)

Appendix  C.10 provides a worksheet to record  the
information  obtained during the performance of
the ESP design procedures.
4.10.1  Data Required
The  data  necessary to  perform the design  steps
consist of the data characteristics previously com-
piled on the HAP Emission Stream Data Forms and
the required HAP control as determined by the ap-
plicable regulations.

In the  case of a permit  review for an ESP, the fol-
lowing data should be  supplied by the applicant:
Reported collection efficiency, %
                                                                        87

-------
Reported drift velocity of particles, ft/sec
Reported collection plate area, ft2

The design criteria and considerations discussed in
this section will be used to evaluate the reasonable-
ness of the applicant's proposed design.
                  Example Case
  Electrostatic precipitation was one of the select-
  ed control techniques for the municipal inciner-
  ator stream. The pertinent data for these proce-
  dures are found on the HAP  Emission Stream
  Data Form (see Figure 3-8).

  Flow rate, Qe/a = 110,000 acfm
  Emission stream temperature, Te  = 400°F
  Particulate content = 3.2 grains/scf - flyash
  Moisture content, Me = 5% (vol)
  HAP content = 10% (mass) cadmium
  Drift velocity of particles, Ud = 0.3 ft/s
  Collection efficiency, CE = 99.9% mass
4.10.2 Pretreatment of the Emission Stream
As discussed  in Section 3.3.1, the temperature of
the emission streams should be within 50° to 100°F
above the stream dew point. Procedures for deter-
mining the  dew point of an emission  stream are
provided in Appendix B.1. If the emission stream
temperature does not fall within the stated range,
pretreatment (i.e., emission stream preheat or cool-
ing)  is necessary.  (Methods  of pretreatment are
briefly discussed in Appendix  B.11, reference  8.)
The primary characteristics affecting ESP sizing are
drift velocity of the particles and flow rate. There-
fore, after selecting a temperature for the emission
stream, the  new stream flow rate must be calculat-
ed. The calculation method depends upon the type
of pretreatment performed; use appropriate stan-
dard industrial equations. The use of pretreatment
mechanical  dust collectors may also be appropri-
ate. If the emission stream contains an appreciable
amount of large particles (20 to 30 (Jim), pretreat-
ment with mechanical dust  collectors  is typically
performed.  (Appendix B.11, reference 8, further de-
scribes the use of mechanical dust collectors.)

4.70.3 ESP Design Variables
Estimating the collection plate area is the important
aspect of sizing an ESP. A secondary consideration
is the material of construction.

4.10.3.1  Collection Plate Area
Although precise specification  of collection plate
area is best left to the vendor, an approximate col-
lection plate area can be calculated using the avail-
able drift velocity value for the gas stream.

As noted earlier, collection plate area is a function
of the emission stream  flow rate, the paniculate
drift velocity, and  desired control efficiency. The
Deutsch-Anderson  equation relates these variables
asfollows:(32,37)
                x ln <1  -CE)
                                  (4.10-1)
where:
AP   =
Qe,a =

Ud   =
CE   =
collection plate area, ft2
emission stream flow rate at actual condi-
tions as it enters the control device, acfm
drift velocity of particles, ft/s
required collection efficiency, decimal frac-
tion
Published data on drift velocities for a number of
industrial applications are presented in Table 4-17.
When unavailable, a drift velocity value for an in-
dustrial application can  be obtained from an ESP
vendor or from literature sources.(30) If no value
for drift velocity is known, 0.30 ft/s for particles of
"average" resistivity (approximately 10 to 2 x 1010
ohm-cm) and 0.10 ft/s for particles having a "high"
resistivity (1011 to 1013 ohm-cm) can be used.(37)
Table 4-17.
    Typical Values for Drift Velocity for Various
    Particulate Matter Applications (37)
               Application
                              Drift Velocity,
                                  ft/s
Pulverized Coal                           0.33 to 0.44
Paper Mills                                 0.25
Open-hearth Furnace                          0.19
Secondary Blast Furnace (80% foundry iron)        0.41
Gypsum                                0.52 to 0.64
Hot Phosphorous                             0.09
Acid Mist (H2SO4)                          0.19 to 0.25
Acid Mist (TiO2)                           0.19 to 0.25
Flash Roaster                                0.25
Multiple-hearth Roaster                        0.26
Portland Cement (wet manufacturing)          0.33 to 0.37
Portland Cement (dry manufacturing)          0.19 to 0.23
Catalyst Dust                                0.25
Gray-iron Cupola (iron-coke ratio = 10)         0.10 to 0.12

Particles with low resistivities impose special de-
sign considerations  on an ESP. Such particles (re-
sistivities from 104 to  107 ohm-cm) are difficult to
collect in an ESP because the particles tend to lose
their charge and drop off the collector plate and
become  reentrained in the gas stream.  In such
cases,  specially designed collecting plates or coat-
ings may be used to reduce reentrainment.(32,37)
Particles with high resistivities also can cause ESP
operating difficulties. High resistivity  particles ac-
cumulate on the collection plates and insulate the
collection plate, thus  reducing the attraction be-
tween  the particles and the  collecting  plate. In
these cases, oversizing an ESP and more frequent
cleaning or  rapping  of the collector plates are nec-
essary. An alternative to a larger ESP  is the use of
conditioning agents to  reduce the resistivity of the
particles.  Consult a  vendor for advice concerning
conditioning agents.
                       88

-------
                 Example Case
  Flow rate, Qe,a = 110,000 acfm
  Drift velocity of particles, Ud = 0.30 ft/s
  Collection efficiency, CE  = 0.999
  From the Deutsch-Anderson equation:

  A  =  -110,000 acfm [In (1 -  0.999)]
   p             60 x 0.30 ft/s

  Ap = 42,200ft2 of collection plate area
4.10.3.2  Materials of Construction
The most common material used in ESP construc-
tion is carbon steel. In cases where the gas stream
contains high concentrations of S03 or where liq-
uid-gas contact areas are involved, stainless steel
may be  required.(30,32,37,39,40) However,  by
keeping  the  emission  stream temperature  above
the dew point and by insulating the ESP (the tem-
perature drop across an insulated ESP should not
exceed 20°F) the use of stainless steel should not
be necessary.


4.10.4 Evaluation of Permit Application
Using Table  4-18, compare the results from this
section and the data supplied by the permit appli-
cant. The calculated values are based on the exam-
ple. In evaluating the reasonableness of ESP design
specifications in a permit application, the main task
will be to examine each parameter in terms of its
compatibility with the gas stream conditions.

If the applicant's collection plate area is less than
the calculated area, the discrepancy will most likely
be the selected drift velocity. Further discussions
with the permit applicant are  recommended to
evaluate the  design assumptions and to reconcile
any apparent discrepancies.
Table 4-18.   Comparison of Calculated Values and Values
            Supplied by the Permit Applicant for ESP's

Drift velocity of particles, Ud
Collection efficiency, CE
Collection plate area, Ap
Calculated Value
(Example Case)a
0.30 ft/s
0.999
42,200 ft2
Reported
Value

"Based on the municipal incinerator emission stream.
4.10.5 Determination of ESP Operating
      Parameters
Many times, optimization  of an  ESP's collection
efficiency occurs in the field after construction. The
following discussion does  not pertain to the pre-
liminary design of an ESP control system, however,
the information presented should be  helpful  in
achieving and  maintaining the desired collection
efficiency for the installed control system.
4.10.5.1 Electric Field Strength
Current in the form of ions from the charging elec-
trodes actually charge the particles. Once the parti-
cles are charged, the electric field strength deter-
mines the amount of charge on the particles.

Field  strength is based  on voltage  and distance
between  the  collecting  plates  and  elec-
trodes.(32,37,39) ESP's are usually operated at the
highest secondary voltage practicable with limited
sparking to maximize collection  efficiency. Spark-
ing represents an instantaneous drop in  voltage,
collapse of the electrostatic field, and momentary
cessation  of paniculate collection. Sparking varies
with the density of the gas stream, material collect-
ed on the electrodes, and humidity and tempera-
ture of the gas stream. When automatic  controls
are used, ESP's  usually  operate  with  a  small
amount of sparking to ensure that the voltage is in
the correct  range and the field  strength  is  maxi-
mized.  Automatic voltage controls  can  control
sparking to a specified  sparking frequency  (typical-
ly 50 to  150 sparks per  minute per section of
ESP).(37) As the spark rate increases, a greater per-
centage of the input power is wasted in the spark
current. Consequently, less useful power is applied
to the collecting electrode.

4.10.5.2 Cleaning Frequency and intensity
Particles  accumulating on  the  collecting plates
must  be removed periodically.  In wet ESP's  the
liquid flowing down the collector surface removes
the particles.(40)  In dry ESP's, the particles are re-
moved by vibrating or rapping the collector plates.
For dry ESP's this is a critical step in the overall
performance because improperly adjusted  or oper-
ating  rappers can cause reentrainment of collected
particles or sparking due to excessive particulate
buildup on the collection plates or discharge elec-
trodes. In  normal operation, dust buildup of 6 to 25
mm is allowed before rapping of a given intensity
is initiated.(32) In this way, collected material falls
off in  large clumps that would not be reentrained. If
rapping is initiated more frequently or if the  inten-
sity of rapping  is lowered, the  resulting  smaller
clumps of particulate matter are more likely to be
reentrained,  reducing  the collection efficiency of
the ESP. Optimal adjustment of the ESP can best be
made by  direct visual inspections  through sight
ports.

4.10.5.3 ESP Collection Efficiency
ESP collection efficiencies less than permitted  can
be the result of operational problems, mechanical
troubles, or improper  design. Typical  operational
problems include improper electrical settings, bad-
ly adjusted rappers, full or nearly full dust hoppers,
and process upsets. Mechanical difficulties typical-
ly are the result of electrode misalignment or ex-
cessive dust buildup on  the electrodes. Basic de-
                                                                        89

-------
sign problems  include  undersized equipment,
reentrainment, or high resistivity particles. The per-
mit applicant  should carefully  examine each  of
these items if the ESP is emitting paniculate emis-
sions from his facility that are in excess of permit-
ted levels.
4.11  Venturi Scrubbers
Venturi scrubbers are designed to serve as a con-
trol device for applications requiring very high col-
lection efficiencies of particles generally between
0.5 to 5.0 (Jim in diameter. They employ gradually
converging and then diverging sections to clean an
incoming gaseous stream. The section connecting
the converging and diverging sections of the scrub-
ber is called the throat. In general, the longer the
throat, the higher the collection efficiency at a giv-
en pressure drop, provided the throat is not so long
that frictional losses become  significant.(42) Typi-
cally, a  liquid  (usually water)  is introduced up-
stream of the throat and flows down the converg-
ing sides into the throat where it is atomized by the
gaseous stream; this method  is called  the "wetted
approach." Alternatively, the liquid can be injected
into the throat  itself by use of nozzles directed at
the throat; this approach is called the  "nonwetted
approach."(42)  The nonwetted  approach works
well when a gas is already  close  to saturation;
however, this method requires that the liquid be
free of particles that could clog the nozzles. Where
inlet gases are  hot and a significant amount of
liquid needs to be evaporated, the wetted approach
is preferred.

Once the liquid is  atomized, it begins  to collect
particles from the gas impacting into the liquid as a
result  of the difference in velocities of the gas
stream and the atomized droplets. As the mixture
decelerates in the expanding section, further im-
paction occurs  causing the droplets to agglomer-
ate. Once the particles have been trapped by the
liquid, a separator  (e.g. cyclone, demisters,  swirl
vanes) can readily  remove the scrubbing  liquid
from the cleaned gas stream.

Appendix C.11 provides a worksheet to record the
information obtained during  the  performance of
the venturi scrubber design procedures.
4.11.1 Data Required
The data necessary to perform the design steps
consist of the HAP emission stream characteristics
previously compiled on the HAP Emission Stream
Data Forms, and the required HAP control as deter-
mined by the applicable regulations.

In the case of a permit review for a venturi scrub-
ber, the following data should be supplied by the
applicant:
Reported pressure drop across venturi, in H20
Performance curve applicable to the venturi
scrubber
Reported collection efficiency, %
                 Example Case
 A venturi scrubber was one of the selected con-
 trol  techniques for  the  municipal incinerator
 emission stream. The pertinent data for these
 procedures are found on the HAP Emission
 Stream Data Form.

 Flow rate Qe/a = 110,000 acfm
 Temperature, Te = 400°F
 Moisture content, Me = 5% vol
 Required collection efficiency, CE = 99.9%
 Particle mean diameter, Dp = 1.0 JJUTI
 Paniculate content = 3.2 grams/scf flyash
 HAP content = 10% (mass) cadmium
4.11.2 Pretreatment of the Emission Stream
As discussed in Section 3.3.1, the temperature of
the emission stream should be within 50° to  100°F
above the stream dew point. Procedures for deter-
mining the dew point of an emission stream  are
provided in Appendix B.1.  If the emission stream
temperature does not fall within the stated range,
pretreatment (i.e., emission stream preheating or
cooling) is necessary. (Methods  of pretreatment
are briefly discussed in reference 32 and Appendix
B.11, references.) If pretreatment is performed, the
emission stream characteristics will be altered. The
primary characteristic affecting  venturi scrubber
design is the saturated gas flow rate (Qe,s), a func-
tion of the emission stream temperature (Te) and
flow rate at actual conditions (Qe,a).(42) Thus, if the
temperature of the emission stream changes, thus
changing the actual flow rate, the saturated gas
flow  rate must be based on the  new actual flow
rate.  The calculation method depends  upon  the
type  of pretreatment performed; use appropriate
standard industrial equations.  The use of pretreat-
ment mechanical dust collectors  may also be  ap-
propriate,  particularly if a "nonwetted" venturi
scrubber is used.

4.11.3 Venturi Scrubber Design Variables
To design a venturi  scrubber, any  one of  three
paths may be chosen: (1) rely on previous experi-
ence with an analogous application, which is best
for plants lacking effluent data; (2) test a scrubber
on the source  itself;  or (3) collect sufficient data
about source stream characteristics, such as parti-
cle size distribution, flow rate,  and temperature, to
utilize existing "performance  curves" for a  given
venturi scrubber. This section is concerned with the
third path. Thus, the most important consideration
becomes the pressure drop across the venturi. A
secondary consideration is materials of construc-
tion.
                      90

-------
4.11.3.1 Venturi Scrubber Pressure Drop
Performance curves are typically logarithmic plots
relating venturi collection efficiency, pressure drop,
and  particle  size.(23,32,43,44) Collection  (control)
efficiency is  usually plotted versus pressure drop
across the venturi (APv) for a particle mean diame-
ter (Dp). Figure 4-20 is a  plot of venturi scrubber
pressure drops for a given collection efficiency and
particle mean diameter for venturi scrubbers man-
ufactured by a specific vendor. Thus, if the particle
mean  diameter for an emission stream and  re-
quired collection  efficiency is known, the pressure
drop across  the venturi can be estimated.  Figure
4-20 is representative of plots likely to be used by
vendors, and does not necessarily  represent char-
acteristics for all venturi scrubbers.

Estimating the pressure drop gives an indication of
whether a venturi  scrubber is  a feasible control
device for a  given stream.  Venturi scrubbers are
                                             used in applications where pressure drops of be-
                                             tween  10 and 80 inches water gauge occur across
                                             the venturi. Venturi scrubbers can operate at pres-
                                             sure drops higher than 80 inches; however, in gen-
                                             eral, a  pressure drop exceeding 80 inches H20 indi-
                                             cates that  a  venturi scrubber  will have difficulty
                                             collecting the particles.(42) Therefore, if the pres-
                                             sure drop  indicated on the performance curve is
                                             greater than  80  inches H20, assume that the ven-
                                             turi scrubber cannot accomplish the desired con-
                                             trol efficiency.


                                             Table 4-19 lists typical pressure drops for venturi
                                             scrubbers for a  variety of applications.  The pres-
                                             sure drops are listed to provide general guidance
                                             for typical values that occur in industry. The values
                                             are not meant to supersede any specific informa-
                                             tion known, and given application may have a pres-
                                             sure drop outside those listed in Table 4-19.
Figure 4-20. Venturi scrubber collection efficiencies.
 o
 E
 o>
 DC
 c
 01
 u
 CD
 Q-
99.9
99.8

99.5
 99

 98
 97
 96
 95

 90


 80
 70

 60

 50

 40
 30

 20

 10
                                                                 rv .X"   Venturi Pressure Drop
                                                                 ^'     (in H20)
           xxx
x  x       x^
                                                  I
                                                     I	I
                                                                                         I
         0.1
                 0.2
         0.3    0.4   0.5  0.6 0.7 0.8 0.9 1.0

           Size of Particles (Aerodynamic Mean Diam.),
                                                                          2.0
                                                                            3.0
4.0  5.0
                                                                         91

-------
Table 4-19.   Pressure Drops for Typical Venturi Scrubber Applications (42)
Application
Boilers
Pulverized coal
Stoker coal
Bark
Combination
Recovery
Incinerators
Sewage sludge
Liquid waste
Solid waste
Municipal
Pathological
Hospital
Kilns
Lime
Soda Ash
Potassium chloride
Coal Processing
Dryers
Crushers
Dryers
General spray
Food spray
Fluid bed
Mining
Crushers
Screens
Transfer points
Pressure drop
(in H20)

15-40
10-12
6-10
10-15
30-40

18-20
50-55

10-20
10-20
10-20

15-25
20-40
30

25
6-20

20-60
20-30
20-30

6-20
6-20
6-20
Application
Iron and Steel
Cupolas
Arc furnaces
BOF's
Sand systems
Coke ovens
Blastfurnaces
Open hearths
Nonferrous metals
Zinc smelters
Copper and brass
smelters
Sinter operations
Aluminum reduction
Phosphorus
Phosphoric acid
Wet process
Furnace grade
Asphalt
Batch plants — dryer
Transfer points
Glass
Container
Plate
Borosilicate
Cement
Wet process kiln
Transfer points
Pressure drop
(in H20)

30-50
30-50
40-60
10
10
20-30
20-30

20-50

20-50
20
50


10-30
40-80

10-15
6-10

25-60
25-60
30-60

10-15
6-12
"Source: Reference 1.
4.11.3.2 Materials of Construction
Proper selection of the materials in constructing a
venturi scrubber ensures long-term operation with
minimal  downtime for repair.  The materials  are
generally chosen based on the corrosive or erosive
nature of the emission stream,  and to a lesser de-
gree, the temperature  of the gas stream. For any
given application, a vendor should be contacted to
ensure correct selection of  materials. A venturi
scrubber will generally be constructed of either car-
bon or stainless steel or a nickel alloy;  it may also
be lined with another material (e.g., ceramics). Ta-
ble 4-20 lists materials of construction  for various
industries and is intended to serve  as a  general
guide  rather than a definitive statement on  the
types of materials used in industry.
                 Example Case
 The required collection efficiency is 99.9% and
 the  particle  mean diameter  in the municipal
 waste incinerator emission stream is estimated
 to be 1.0 n,m; therefore:
 APV = 47 in H20 (Figure 4-20)
 Since the estimated venturi pressure drop value
 of APV is not greater than 80 in H20, this venturi
 scrubber should be able to accomplish the de-
 sired control efficiency. Table 4-20 indicates the
 venturi scrubber should  be constructed of 316L
 stainless steel.
4. 1 1.4 Sizing of Venturi Scrubbers
If a venturi scrubber is found to be a feasible con-
trol choice for a given emission stream, it is then
sized. Venturi scrubbers can be sized using either
the flowrate at inlet conditions (Qe,a) or the saturat-
ed gas flowrate (Qe,s).(45) Vendors  may use either
parameter; the cost data presented in Chapter 5 are
based on Qea. However, more current cost curves
based on Qes may be  available;  therefore, Qes
should be calculated. A psychrometric chart (Figure
4-21) can be used to determine the saturated gas
temperature (Tes), and Qes can then be calculated
using the following formula:
Qe,s  =  Qe,a x (T
               e,s
                    460)/(Te + 460)
                                        (4.11-1)
where:
Qe/s  =  saturated emission stream flow rate, acfm
     =  temperature  of  the  saturated  emission
        stream, °F
 1 e,s
4.11.5  Evaluation of Permit Application
Using  Table 4-21, compare the results of this sec-
tion and the data supplied by the permit applicant.
The calculated values in the table are based on the
example. Compare the estimated APV and the re-
ported pressure drop  across the venturi, as sup-
plied by the permit applicant.
                      92

-------
Table 4-20.    Construction Materials for Typical Venturi Scrubber Applications"
Application
Boilers
Pulverized coal
Stoker coal
Bark
Combination
Recovery
Incinerators
Sewage sludge
Liquid waste
Solid waste
Municipal
Pathological
Hospital
Kilns
Lime
Soda Ash
Potassium chloride
Coal Processing
Dryers
Crushers
Dryers
General spray dryer
Food spray dryer
Fluid bed dryer
Mining
Crushers
Screens
Transfer points
Construction
Material

31 6L stainless steel
31 6L stainless steel
Carbon steel
31 6L stainless steel
Carbon steel or 31 6L stainless steel

316L stainless steel
High nickel alloy

316L stainless steel
31 6L stainless steel
High nickel alloy

Carbon steel or stainless steel
Carbon steel or stainless steel
Carbon steel or stainless steel

304 stainless steel or 31 6L stainless steel
Carbon steel

Carbon steel or stainless steel
Food-grade stainless steel
Carbon steel or stainless steel

Carbon steel
Carbon steel
Carbon steel
Application
Iron and Steel
Cupolas
Arc furnaces
BOF's
Sand systems
Coke ovens
Blast furnaces
Open hearths
Nonferrous Metals
Zinc smelters
Copper and brass
smelters
Sinter operations
Aluminum reduction
Phosphorus
Phosphoric acid
Wet process
Furnace grade
Asphalt
Batch plants — dryer
Transfer points
Glass
Container
Plate
Borosilicate
Cement
Wet process kiln
Transfer points
Construction
Material

304-31 6L stainless steel
31 6L stainless steel
Carbon steel (ceramic lined)
Carbon steel
Carbon steel
Carbon steel (ceramic lined)
Carbon steel (ceramic lined)

Stainless steel or high nickel

Stainless steel or high nickel
Stainless steel or high nickel
High nickel


316L stainless steel
31 6L stainless steel

Stainless steel
Carbon steel

Stainless steel
Stainless steel
Stainless steel

Carbon steel or stainless steel
Carbon steel
                  Example Case
 Determine saturated gas flow rate:
 Emission stream flow rate, Qe,a = 110,000 acfm
 Moisture content, Me = 5% vol
 Emission stream temperature, Te = 400°F

 Convert Me to units of Ib H20/lb dry air, decimal
 fraction:
 (Me/100) (18/29) = (5/100) (18/29) =
 0.031 Ib H20/lb dry air
 Tes = 127°F (Figure 4-21)
 Qe,s =  (110,000) x
 75,000 acfm
(127 +  460)/(400  + 460) =
                                                    Table 4-21.
                                            Comparison of Calculated Values and Values
                                            Supplied by the Permit Applicant for Venturi
                                            Scrubbers
                                                           Calculated Value
                                                           (Example Case)3
Reported
 Value
                                Particle mean diameter, Dp          1.0 n,m
                                Collection efficiency, CE              0.999
                                Pressure drop across venturi, APV    47 in H20
                                "Based on the municipal incinerator emission stream.
If the estimated and reported values differ, the dif-
ferences may be due to the applicant's use of an-
other performance chart, or a discrepancy between
the required and reported collection  efficiencies.
Discuss the details of the  design and operation of
the system with the applicant. If there are no differ-
ences between the estimated and reported values
for APV, the design and operation of the system can
be considered appropriate based on the  assump-
tions employed in this handbook.
                                                                           93

-------
Figure 4-21.  Psychrometric chart, temperature range 0° - 500°F, 29.92 in Hg pressure.
  0
              50
                       100      150
200       250       300

  Temperature, °F
                                                                  350
                                   400
4.12 References

  1. U.S. EPA. Organic  Chemical Manufacturing.
    Volume 4: Combustion Control Devices. EPA-
    450/3-80-026. December 1980.

  2. U.S. EPA. Reactor Processes in  Synthetic Or-
    ganic Chemical Manufacturing Industry —
    Background Information  for Proposed Stan-
    dards. Draft EIS. October 1984.

  3. Hougen,  O.A., K. M. Watson, and R. A. Ragatz.
    Chemical Process Principles Part I:  Material
    and Energy Balances. Asia Publishing House.
    Bombay. 1962.

  4. U.S. EPA. Parametric Evaluation of VOC/HAP
    Destruction  Via Catalytic Incineration.  EPA-
    600/2-85-041. April 1985.

  5. U.S. EPA. Control of Volatile Organic Emis-
    sions from Existing Stationary Sources — Vol-
    ume I: Control Methods  for Surface Coating
    Operations.   EPA-450/2-76-028.  November
    1976.
          6. U.S. EPA. Guideline Series. Control of Volatile
             Organic Compound Emissions From Manufac-
             ture of High-Density Polyethylene, Polypropy-
             lene, and Polystyrene Resins.  EPA-450/3-83-
             008. November 1983.

          7. U.S. EPA. Afterburner Systems Study. EPA-R2-
             72-062. August 1972.

          8. U.S. EPA. Evaluation of Control Technologies
             for  Hazardous Air Pollutants — Appendices.
             EPA-600/7-86-009b (NTIS PB 86-167/038/AS).
             October 1985.

          9. Federal Register. Volume 50. April 16,1985. pp.
             14941-14945.

          10. U.S. EPA. Evaluation of the Efficiency of Indus-
             trial Flares: Test  Results.  EPA-600/2-84-095.
             May 1984.

          11. Chandrasekhar, R., and E. Poulin.  Control of
             Hydrocarbon Emissions From Cotton and Syn-
             thetic Textile Finishing Plants.  EPA  Contract
             No. 68-02-3134. May 1983.
                      94

-------
12.  U.S.  EPA.  Organic Chemical Manufacturing
    Volume 5: Adsorption, Condensation, and Ab-
    sorption Devices. EPA-450/3-80-027. December
    1980.

13.  Parmele, C. S., W. L O'Connell, and H. S. Bas-
    dekis. Vapor-phase Adsorption Cuts Pollution,
    Recovers Solvents. Chemical Engineering. De-
    cember 31, 1979. pp. 58-70.

14.  Calgon  Corporation, Pittsburgh, Pennsylvania.
    In-house data.

15.  Tomany, J. P. Air Pollution: The Emissions, the
    Regulations, and the Controls. American Else-
    vier Publishing Company, Inc. New York.

16.  Manzone, R. R., and D. W.  Oakes. Profitably
    Recycling  Solvents From Process Systems.
    Pollution Engineering. October 1973. pp.  23-24.

17.  Vatavuk, M. W., and R. B.  Neveril. Part XIV:
    Costs of Carbon Adsorbers. Chemical  Engi-
    neering. January 24, 1983. pp. 131-132.

18.  Chemical Engineer's  Handbook.  Perry,  R. H.,
    and C. H. Chilton, eds.  Fifth  Edition. McGraw-
    Hill Book Company. New York. 1973.

19.  Smith, J. M., and H. C. Van  Ness. Introduction
    to Chemical Engineering Thermodynamics.
    Second Edition. McGraw-Hill Book Company,
    Inc. and Kogakusha Company, Ltd. Tokyo.
    1959.

20.  Treybal, R. E. Mass-Transfer Operations. Third
    Edition. McGraw-Hill Book Company.  New
    York. 1980.

21.  Buonicore, A. J., and L.  Theodore. Industrial
    Control Equipment for Gaseous Pollutants.
    Volume I.  CRC Press, Inc.  Cleveland,  Ohio.
    1975.

22.  Kohl, A., and F. Riesenfeld. Gas Purification.
    Second  Edition.  Gulf  Publishing  Company.
    Houston, Texas. 1974.

23.  U.S. EPA. Wet Scrubber System Study, Volume
    I:  Scrubber Handbook.  EPA-R2-72-118a. Au-
    gust 1972.

24.  Vatavuk, W.  M., and R.  B.  Neveril. Part XIII.
    Costs of Gas Absorbers. Chemical Engineering.
    October 4,  1982. pp. 135-136.

25.  Kern, D. Q. Process Heat Transfer. McGraw-Hill
    Book Company, Inc. and Kogakusha Company,
    Ltd. Tokyo. 1950.

26.  Ludwig, E. E. Volume III.  Applied Process De-
    sign  for Chemical and Petrochemical Plants.
    Gulf  Publishing Company.  Houston, Texas.
    1965.
27. Lange's Handbook of Chemistry. Dean, J. A.,
    ed. Twelfth Edition. McGraw-Hill  Book Com-
    pany. New York. 1979.

28. Siebert, P. C. Handbook on Fabric Filtration. IIT
    Research Institute. Chicago. April 1977.

29. U.S. EPA. Handbook of Fabric Filter Technol-
    ogy, Volume I:  Fabric Filter  Systems Study.
    APTD 0690. December 1970.

30. U.S. EPACapital and Operating Costs of Se-
    lected Air Polution Control Systems. EPA-450/
    5-80-002. December 1978.

31. The Fabric Filter Manual. The Mcllvaine Com-
    pany. Northbrook, Illinois. 1975.

32. Environmental Engineers' Handbook, Volume
    II: Air Pollution.  Liptak, B. G., ed. Chilton Book
    Company. Radnor, Pennsylvania. 1974.

33. U.S. EPA. Control Techniques for Particulate
    Emissions from Stationary Sources — Volume
    2. EPA-450/3-81-005b. September 1982.

34. U.S. EPA. Control Techniques for Particulate
    Emissions From Stationary Sources — Volume
    I. EPA-450/3-81-005a. September 1982.

35. Strauss, W. Industrial Gas Cleaning, 2nd Edi-
    tion. Pergamon Press, Oxford, England. 1975.

36. U.S. EPA. Procedures Manual for Fabric Filter
    Evaluation. EPA-600/7-78-113. June 1978.

37. U.S. EPA. Air Pollution Engineering Manual.
    AP-40. May 1973.

38. U.S. EPA. Particulate Control Highlights:  Re-
    search on Fabric Filtration  Technology. EPA-
    600/8-78-005d. June 1978.

39. U.S. EPA. A Manual of Electrostatic Precipitator
    Technology, Part 1 —  Fundamentals.  APTD
    0610. 1970.

40. Perry's Chemical Engineers' Handbook.  Perry,
    R.  H., and D.  Green, eds. Sixth Edition.
    McGraw-Hill Book Company. New York. 1984.

41. The Electrostatic Precipitator Manual. The Mcll-
    vaine Company. Northbrook, Illinois. 1975.

42. Air Pollution Control and Design  Handbook:
    Part 2. Cheremisinoff, P. N., and R. A. Young,
    eds. Marcel Dekker, Inc. New York. 1977.

43. U.S. EPA. Wet Scrubber Performance Model.
    EPA 600/2-77-127. August 1977.

44. U.S. EPA. TI-59 Programmable Calculator Pro-
    grams for Opacity, Venturi Scrubbers and Elec-
    trostatic  Precipitators. EPA-600/8-80-024. May
    1980.

45. U.S. EPA. The Cost Digest. Cost Summaries of
    Selected Environmental Control Technologies.
    EPA-600/8-84-010. October 1984.
                    95

-------
                                            Chapter 5

                                  Cost Estimation Procedure
 5.1  Objective
 This chapter provides generalized procedures for
 estimating capital and annualized costs (June 1985
 dollars) for a given add-on HAP control  system.
 (Note: Calculation of the cost of HAP waste dispos-
 al is outside the scope of this handbook; however,
 this cost must be included in any rigorous control
 cost estimation.) The procedures are presented in a
 step-by-step  format and  illustrated  at each step
 with cost calculations pertaining to the thermal in-
 cinerator system example discussed in Section 4.2.
 Blank standard cost calculation worksheets are pro-
 vided in Appendix C.12.

 Only the process HAP control systems presented in
 Chapter 4 are discussed in this chapter. The cost of
 fugitive emission controls are outside the scope of
 this  handbook; however, an EPA report  entitled
 Identification, Assessment, and Control of Fugitive
 Paniculate Emissions (1) can be used for estimat-
 ing costs of fugitive emission controls.

 5.2  Total Capital Cost
 In this handbook the total capital cost includes only
 manufacturing area costs; therefore,  it excludes
 offsite costs.  The total capital cost of a control sys-
 tem is the  sum of direct  costs, indirect costs, and
 contingency  costs.  Direct costs include the total
 purchased  equipment cost (i.e., the  major equip-
 ment purchased cost plus the auxiliary equipment
 purchased cost), instrumentation and controls,
 freight and taxes, and installation costs (i.e.,  foun-
 dation and supports, erection and handling, electri-
 cal, piping, insulation, and  painting). (Note: The
 summation of the total purchased equipment cost,
 the cost of  instrumentation  and controls, and
 freight and taxes is  defined as the total purchased
 cost.) Indirect costs  consist of inhouse engineering
 design and supervision costs, architect and  engi-
 neering contractor expenses, contractor fees, con-
 struction expenses,  and preliminary testing costs.
 An example of contingency costs are penalties in-
 curred for failure to  meet completion  dates or per-
formance specifications.

The capital cost estimation procedure presented  in
this handbook is for a "factored" or  "study" esti-
 mate. Usual reliability for a study type estimate  is
 ± 30 percent. To determine the total capital cost by
 a factored cost estimate, a reliable estimate of the
total purchased cost is calculated  and predeter-
mined factors are applied to determine all  other
capital cost elements. Therefore, the procedure to
estimate the total capital cost is as follows: (1)
obtain the total purchased equipment cost by esti-
mating the purchased cost of major and auxiliary
equipment; (2) estimate the cost of instrumenta-
tion and controls plus freight and taxes as a per-
centage of the total purchased equipment cost; (3)
estimate the total purchased cost by  adding (1) and
(2) above; and (4) estimate total capital cost by
applying a predetermined cost factor to the total
purchased cost.

5.2.1 Estimation of Major Equipment Purchased
     Cost
The major equipment purchased cost (i.e., the cost
of the major components that comprise the control
system)  is related to a specific equipment design
parameter and can be expressed either analytically
or graphically. Table 5-1 presents a list of the de-
sign  parameters needed for costing the HAP  con-
trol equipment, and it identifies the figure that pre-
sents  the applicable  purchased  cost  curve.
Gathering current costs from vendors was beyond
the scope of this project and thus, necessitated use
of dated cost data compiled by others. In general,
Table 5-1.   Identification of Design Parameters and
          Cost Curves for Major Equipment
Control
Equipment
Thermal Incinerator
Heat Exchanger
Catalytic Incinerator
Carbon Adsorber

Absorber

Condenser

Fabric Filter
Electrostatic Precipitator
Venturi Scrubber


Design
Parameter3
V0
A
Of.
creq

wtool
Dcol
A
"con
Ref
A
"nc
AP
Qa,a


Cost Curve
Figure No.
5-1
5-2
5-3
5-4b
5-5°
5-6
5-7
5-8
5-9
5-10
5-11
5-12
5-13
5-14
aSee Nomenclature for definitions of variables.
bPackaged carbon adsorbers.
cCustom carbon adsorbers.
                                               97

-------
cost estimates should not be escalated beyond 5
years. If more recent cost data are available, they
should be substituted for the cost curve data pre-
sented. These cost curves should not be extrapolat-
ed beyond their range. The cost data presented in
these figures were obtained from cost information
published in EPA reports. (2,3) To escalate the cost
data to June 1985 dollars, multiply the cost esti-
mate by the ratio of the Chemical Engineering Fab-
ricated Equipment (FE) cost indices for June 1985
and the date of the cost data. For example, if a cost
is given in December 1977 dollars, it is converted to
June 1985 dollars using a factor of 1.49 (336.0
[June 19851/226.2 [Dec. 1977]). Table 5-2 presents
the monthly FE cost indices from December 1977
through June 1986.
Using the specific value for the design variable,
obtain purchased costs from the specific cost curve
for each major control system component. Present-
ed below are brief descriptions of the equipment
costs included in each HAP control cost curve.
The cost curve for thermal incinerators (Figure 5-1)
includes the fan plus instrumentation and control
costs, in addition to the major equipment pur-
chased cost. If the HAP control system includes a
heat exchanger, its cost (Figure 5-2) is part of the
major equipment purchased cost and thus, must be
added. The remaining auxiliary equipment
(ductwork and stack) purchased costs and costs of
freight and taxes must be added to obtain the total
purchased cost.
The cost curve for catalytic incinerators (Figure 5-3)
provides the cost of an incinerator less catalyst.
Catalyst costs (Table 5-3) and the cost of a heat
exchanger, if applicable, (Fig. 5-2) must be added to
obtain the major equipment purchased cost. All
auxiliary equipment (ductwork, fan, and stack) pur-
chased costs, the cost of instrumentation and con-
trols, and freight and taxes must be added to obtain
the total purchased cost.

Figure 5-1. Prices for thermal incinerators, including fan and
motor, and instrumentation and controls costs. (2)
110


l/> 100
CO
"o
•Q
£ qn
c^ yu
IN

°- 80
vt
- 7n
ft /U
O
(0
1 fin
C DU
'o
c
Rfl



^^,


^r
jr
/
/
/
/





i i i i i i
0 200 400 600 800 1,000 1,200
Combustion Chamber Volume (ft3)
Table 5-2. Chemical Engineering Fabricated Equipment (FE) Cost Indices'
Date FE Date FE
Dec. 1977 226.2 Jan. 1980 273.8
Feb. 1980 276.9
Jan. 1978 226.6 Mar. 1980 277.7
Feb. 1978 233.0 Apr. 1980 289.3
Mar. 1978 233.6 May 1980 290.9
Apr. 1978 237.1 June 1980 291.3
May 1978 237.3 July 1980 296.7
June 1978 237.4 Aug. 1980 297.3
July 1978 238.6 Sept. 1980 298.1
Aug. 1978 243.3 Oct. 1980 301.2
Sept. 1978 243.2 Nov. 1980 302.5
Oct. 1978 243.8 Dec. 1980 304.0
Nov. 1978 244.1
Dec. 1978 245.2 Jan. 1981 305.9
Feb. 1981 307.1
Jan. 1979 245.2 Mar. 1981 314.7
Feb. 1979 252.5 Apr. 1981 321.9
Mar. 1979 253.1 May 1981 321.6
Apr. 1979 253.7 June 1981 322.9
May 1979 258.3 July 1981 325.6
June 1979 259.9 Aug. 1981 325.7
July 1979 262.6 Sept. 1981 326.7
Aug. 1979 264.2 Oct. 1981 330.8
Sept. 1979 266.6 Nov. 1981 329.4
Oct. 1979 271.6 Dec. 1981 328.9
Nov. 1979 272.6
Dec. 1979 273.7 Jan. 1982 324.5
Feb. 1982 323.4
Date FE Date FE
Mar. 1982 324.1 May 1984 334.6
Apr. 1982 327.8 June 1984 333.8
May 1982 329.1 July 1984 335.4
June 1982 327.5 Aug. 1984 335.1
July 1982 327.1 Sept. 1984 335.9
Aug. 1982 326.2 Oct. 1984 335.0
Sept. 1982 326.7 Nov. 1984 335.4
Oct. 1982 325.8 Dec. 1984 336.5
Nov. 1982 324.8
Dec. 1982 325.1 Jan. 1985 336.9
Feb. 1985 336.5
Jan. 1983 324.4 Mar. 1985 336.6
Feb. 1983 327.6 Apr. 1985 338.0
Mar. 1983 326.8 May 1985 336.0
Apr. 1983 326.6 June 1985 336.2
May 1983 327.1 July 1985 336.4
June 1983 327.3 Aug. 1985 335.3
July 1983 327.0 Sept. 1985 335.9
Aug. 1983 327.1 Oct. 1985 335.4
Sept. 1983 328.0 Nov. 1985 335.7
Oct. 1983 327.8 Dec. 1985 336.8
Nov. 1983 328.9
Dec. 1983 330.1 Jan. 1986 332.5
Feb. 1986 319.2
Jan. 1984 331.5 Mar. 1986 317.0
Feb. 1984 333.0 Apr. 1986 310.6
Mar. 1984 332.9 May. 1986 310.9
Apr. 1984 333.8 June 1986b 310.8
aSource: Chemical Engineering, McGraw-Hill Publications.
bPreliminary.

                          98

-------
Figure 5-2.   Prices for thermal oxidation recuperative heat
           exchangers. (2)
J52.000-
"5
ro1
o
o
o
000-
800--
600-^
400--

200--
   100-
0)
01
C
CO
.C
(J
X
LU
 80-
 60--
    20-
         200  500  1,000          10,000        100,000

         Recuperative Heat Exchanger Surface Area (ft2)
                                                                  Example Case
                                                 The example thermal incinerator system case
                                                 (see Section 4.2) consists of an incinerator with a
                                                 combustion chamber volume  (Vc)  of approxi-
                                                 mately 860 ft3  and  a primary heat exchanger
                                                 with a surface area  (A) of approximately 4,200
                                                 ft2. From the cost data presented in Figures 5-1
                                                 and 5-2, June 1985 cost estimates are obtained
                                                 as follows:

                                                 (a) Incinerator plus instrumentation and control
                                                     costs—
                                                     $98,000 x (336.2/226.2) = $145,700
                                                     (Note: 12/77 dollars escalated to reflect 6/85
                                                     dollars.)

                                                 (b) Heat exchanger cost—
                                                     $85,000 x (336.2/273.7) = $104,400
                                                     (Note: 12/79 dollars escalated to reflect 6/85
                                                     dollars.)
 Figure 5-3.   Prices for catalytic incinerators, less catalyst.{4)

  1,000
jo 500
"o
-o
O)
C:
CM
o
o

I 100
    50
 o
 S
 CD
 C
 'o
 C
    1051
1 1 1 1
1 1 1
1 1 1 1
1 1 1
MM
.00  1,000
                   5,000 10,000
                                           100,000
                                       50,000
            Emission Stream Flow Rate, Qe (scfm)
Table 5-3.
Chemical
     Unit Costs for Various Materials
     (6/85 dollars)
                                          Cost
Refrigerant
(ethylene glycol)
Activated Carbon
Catalyst (platinum-based)
                                $0.31/lb (8)

                                $1.92/lb(9)
                                $2,750/ft3(10)
                                                Two cost  curves  are  presented for carbon  ad-
                                                sorbers: Figure 5-4 for packaged carbon adsorbers
                                                and Figure 5-5  for custom carbon adsorbers.  The
                                                cost curve  for packaged carbon adsorbers includes
                                                the fan plus  instrumentation and control costs, in
                                                addition to the major equipment purchased cost.
                                                The  cost  of  the  remaining auxiliary  equipment
                                                (ductwork  and  stack), as well as, costs of freight
                                                and taxes  must be added to obtain the total pur-
                                                chased cost.  The cost curve for custom carbon ad-
                                                sorbers does not include the cost of carbon (part of
                                                the major  equipment purchased cost), however, it
                                                does include the cost of instrumentation and con-
                                                trols; the cost of carbon is obtained from Table 5-3.
                                                Figure 5-4.  Prices for carbon adsorber packages.(5)
u>
jo  90
"o

E  80
CM
C.  70
                                                        60

                                                        50

                                                        40
                                                      CD
                                                      D)
                                                      3
                                                      o
                                                     30

                                                     20

                                                     10
                                                              j	i	i	i	i
                                                        01    2345678!

                                                                 Carbon Weight, Creq (1,000 Ib)

                                                        Price includes carbon, beds, fan and motor, and
                                                        instrumentation and controls.
                                              10
                                                                            99

-------
Figure 5-5.  Prices for custom carbon adsorbers, less
          carbon.(5)
                                                    Figure 5-7.   Prices for absorber platform and ladders.(6)
-S  800
o
73
1^
C;
CN
O
O
O
Q.
£
o
0)
d
o
.a
2
   700

   600

   500

   400

   300

   200

   100
      0    20   40   60   80  100  120  140  160  180  200

               Carbon Weight, Creq (1,000 Ib)
      Price includes beds, instrumentation and controls,
      and a steam regenerator.
                                                     o
                                                     13


                                                     I
                                                    T3
                                                    TJ
                                                    JS
                                                    T3
                                                     C
                                                     CD
                                                     O

                                                     JO
                                                     Q.
                                                                                 Hc = 40 ft
                                                                                 Hc = 33 ft
                                                                                 Hc = 27 ft
                                                                    Column Diameter, DCO|Umn (ft)
All auxiliary equipment (ductwork, fan, and stack)
purchased costs and freight and taxes must be add-
ed to obtain the total purchased cost.
The cost curve for absorbers (Figure 5-6) does not
include the cost of packing, platforms, and ladders.
The cost of platform and ladders (Figure 5-7) and
packing (Table 5-4) must  be added  to obtain the
major equipment purchased cost. All auxiliary
equipment (ductwork, fan,  and stack) purchased
costs, the cost of instrumentation and controls, and
freight and taxes must be added to obtain the total
purchased cost.
Figure 5-6.   Prices for absorber columns.(6)

  1,000 (=	
_  500
2?
o
T3


1
O
O
o
0)
.a
   100
    50
    10
                       i  I i M 11
                                  i   i  i  I  i i 111
      0  5    10          50  100         500  1,000

             Column Weight, Wtcolumn (1,000 Ib)

       Price includes manholes, skirts, and painting.
                                                    Table 5-4.    Price of Packing for Absorber Systems (6)
Packing Type and
Material
Packing Diameter, inches
Pall Rings
Carbon steel
Stainless steel
Polypropylene
Berl Saddles
Stoneware
Porcelain
Intalox Saddles
Polypropylene
Porcelain
Stoneware
Packing Rings
Carbon steel
Porcelain
Stainless steel
Cost/Ft3
(6/81 Dollars)
1
24.3
92.1
21.9

28.1
34.5

21.9
19.4
18.2
30.3
13.2
109.0
1.5
16.5
70.3
14.8

21.7
25.6

14.8
13.3
19.8
10.6
82.6
2
15.1
0.8
13.8

—
—

13.6
13.3
12.2
17.0
9.7
22.9
3
_
—
—

—
—

7.0
12.2
11.0
13.9
8.1
—
                                                    The cost curve for condensers (Figure 5-8) yields
                                                    the total capital cost for cold water condenser sys-
                                                    tems. For  systems needing refrigerant  (ethylene
                                                    glycol), the applicable cost from Figure 5-9 must be
                                                    added to the cost obtained from Figure 5-8. Since a
                                                    total capital cost is determined, no additional cost
                                                    estimates are necessary; therefore, proceed to Sec-
                                                    tion 5.3 to calculate annualized operating costs.

                                                    The cost curve for a negative pressure fabric filter
                                                    (Figure 5-10) does not include  the the cost of bags
                                                    (Table 5-5), which depend upon type of fabric used.
                                                    This cost must be added to obtain the major equip-
                                                    ment purchased cost. All  auxiliary equipment
                                                    (ductwork, fan, and stack) purchased costs, the cost
                       100

-------
 Figure 5-8.   Total capital costs for cold water condenser "
            systems.(7)

    5,000
  "o
  •a
(fl  -
<38
— o
to  -
•
   « 1,000
   CM

   5  500
3 I
II
   Ul
   c
    20,000 ft2
Pulse-jet"
Mechanical Shaker
Reverse Air
Dacron
0.40

0.35

0.60
0.25
0.25
Orion
0.65

0.50

0.95
0.35
0.35
Nylon
0.75

0.70

—
0.45
0.45
Nomex
1.15

1.05

1.30
0.65
0.65
Glass
0.50

0.45

—
0.30
0.30
Polypropylene
0.65

0.55

0.70
0.35
0.35
Cotton
0.45

0.40

—
0.40
0.40
 'For heavy felt, multiply source by 1.5.
                                                                                 707

-------
Figure 5-11.  Prices for insulated electrostatic precipitators.(2)

  10,000
   5,000
c;
CN
   1,000
    500
    100
                     i  i  I i i i i I
i  I i
            10          50   100          500  1,000

              Collection Plate Area, Ap (1,000 ft2)
             Figure 5-13. Required steel thicknesses for venturi
                       scrubbers.(2)
                                                           2        5     10             50     100

                                                                Emission Stream Flowrate, Qe,a (1,000 acfm)

                                                            For use with Figure 5-12,
                                                                                                      200
The cost curve for venturi scrubbers (Figure 5-12)
includes the cost of instrumentation and controls,
in addition to the major equipment purchased cost.
This cost curve is based on a venturi scrubber con-
structed from 1/8-inch carbon steel. Figure 5-13 is
used to determine if 1/8-inch steel is appropriate for
a given application  (use the higher curve). If thicker
steel is required. Figure 5-14 yields a price adjust-
ment factor for various steel thicknesses; this fac-
tor is used  to escalate the cost obtained from Fig-
ure 5-12. In addition, if stainless steel is required
(see Section  4.11.3.2)  multiply the scrubber cost
estimate by 2.3 for 304L stainless steel or by 3.2 for
316L stainless  steel.  Costs of all  auxiliary  equip-
ment (ductwork, fan,  and stack) and  freight and
taxes must be added to obtain the total purchased
cost.
Figure 5-12.  Prices for venturi scrubbers.(2)
    60
    50
    40
    30
0.


-------
5.2.2.1 Ductwork Purchase Cost
The ductwork purchase cost is typically proportion-
al to the ductwork weight, which is a function of: (1)
the material of construction, (2) length, (3) diame-
ter, and (4) thickness. Carbon steel ducts are nor-
mally used for noncorrosive flue gases at tempera-
tures below 1,150°F.  Stainless  steel  ducts are
generally used  with  gas temperatures between
1,1 SOT to 1,500°F, or if the gas  stream contains
corrosive materials. Figures 5-15 and  5-16 present
purchase costs for carbon steel and stainless steel
ducts, respectively.  It is  assumed that the major
portion of ductwork  is utilized to transport the
emission  stream from the process to the control
system;  therefore,  the flow rate  of the emission
stream  at actual conditions  (Qe,a) is  used to size
the ductwork.
                                                   Table 5-6.    Identification of Design Parameters and
                                                              Cost Curves for Auxiliary Equipment
                                                   Auxiliary Equipment
                                                                     Design Parameters
                                        Cost Curve
                                        Figure No.
                                                   Ductwork
                                                   Fan8
                                                   Stack
                                                                   Diameter                 5-15
                                                                   Length                   5-16
                                                                   Material of construction
                                                                   Actual air flow rate          5-17
                                                                   Pressure drop
                                                                   Gas stream velocity
                                                                   Diameter                 5-18
                                                                   Length                   5-19
                                                                   Material of construction
                                                   a Assumed to be located downstream of the control system and
                                                    ductwork.
                                                   Figure 5-15. Carbon steel straight duct fabrication price at
                                                             various thicknesses.(2)
Without specific information, assume the following
items to simplify the costing procedure:
(1)  The  ductwork  is  constructed with  3/16-inch
    thick plate.
(2)  The duct length equals 100 feet.
(3)  The duct diameter is calculated  using a duct
    gas velocity of 2,000 ft/min. Therefore:
D
    duct
       =  1 2  -
                   i9*  = 0.3028 (Qe a)     (5-1 )
                   uduct /
    where:
    Dduct  - duct diameter, in
    Qea   = emission stream flow rate of actual
            conditions, acfm
    Uduct  - velocity of gas stream in duct, ft/min
                  Example Case
 In the example case,  since no specific data  on
 the ductwork are available, use the above as-
 sumptions to cost the ductwork. The duct diam-
 eter is calculated according to item (3) above.
 The emission stream  flow rate at  actual condi-
 tions is  approximately 16,500 acfm; therefore,
 the duct diameter equals 39 inches.

         0.3028 (16,500)* -
D
   duct
                        39 in
 The length of the ductwork is assumed to be 100
 feet of 3/16-inch thick plate. Since the emission
 stream  treated contains no chlorine or  sulfur
 compounds  (i.e.,  it is a noncorrosive emission
 stream) and the gas temperature is 960°F, car-
 bon steel ductwork is used. From Figure 5-15,
 the cost of the ductwork is estimated as follows:

 $52/ft x 100 ft x (336.2/226.2) = $7,700

 (Note: 12/77 dollars escalated to 6/85 dollars.)
                                                                40
                                                                     80      120     160

                                                                    Duct Diameter, Dd (in)
                                                                                              200
                                                    Figure 5-16.  Stainless steel straight duct fabrication price at
                                                               various thicknesses.(2)
                                                      1,600


                                                      1,400


                                                      1,200


                                                      1,000
CM
~   800

c

g>   600
                                                       400
                                                   o
                                                       200
                                                          20    40    60   80   100    120

                                                                      Duct Diameter,Dd (in)
                                                                                        140   160
                                                                          103

-------
5.2.2.2  Fan Purchase Cost
The fan purchase cost (Figure 5-17) is a function of
the flow rate moved by the fan and the pressure
drop  (AP)  across the control system. The fan is
assumed to be  located downstream  of the final
control device in the control system. Therefore, the
fan capacity  must be  based on  the final  control
device's exit gas flow rate  at  actual  conditions
(Qfg,a).  Control  system pressure drop (AP) is  the
total of the pressure drops across the various con-
trol system equipment,  including  the stack and
ductwork. Table 5-7 presents conservative pressure
drops across specific control system components
which can  be used if specific data are not available.
   Using the actual flow rate and total AP parameters,
   obtain the  fan purchased cost from  Figure 5-17.
   Fans are categorized  into Classes I to IV according
   to  control  system  pressure drop. Guidelines are
   presented in Figure 5-17 to determine which class
   of fan to use.  There is some overlap between the
   classes. The lower class fan is generally selected
   due to cost savings. To estimate the cost of a motor
   for the  fan, multiply the fan cost by 15 percent.
   (Note: The  fan and motor costs are included in the
   cost curves for thermal incinerators and  packaged
   carbon adsorbers.)
Figure 5-17.  Fan prices.(2)
  1,000,000 r
E  100,000 :
CD
DC
§
    10,000 :
    1,0000
     Fan Wheel Diameter
           98 in

           80 in

            60 in
                                                                                         q 100,000
                                                                                         - 10,000
                                                                                                   o
                                                                                                  T3
                                                                                         - 1,000    <5
                                                                                           100
                                                          Performance Range8
                                                             Single Width
         1    2    4  6 8 10  20 30

            Pressure Drop, AP (in H2O)
5 in @ 2,300 fpm to 2-1/2 in @ 3,200 fpm
S-Vi in @ 3,000 fpm to 4-1/4 in @ 4,175 fpm
13-V2 in @ 3,780 fpm to G-% in @ 5,260 fpm
Above Class III specifications 	
                                        "Performance Range designations are indicated by static pres-
                                        sure (inches of water) at fan outlet velocity (feet per minute).

                                        Notes:
                                        Fan price is lower at higher pressure drop because smaller fan
                                        wheel is used at higher rpm.
                                        For high temperature environment add 3% (250°F to 600°F).
                                        For stainless steel construction multiply price by 2.5.
                       704

-------
Figure 5-18.  Carbon steel stack fabrication price for 1/4 in
           plate.(2)
o
•o
I*.
s
tf>
 
-------
(3) The stack diameter is calculated using a stack
   exit velocity of 4000 ft/min. Therefore:
   D
    stack
= 12  -x
                    stack
= 0.2141(0^)*  (5-2)
   where:
   Dstack  - stack diameter, in
   Qfg,a   = flue gas  flow rate at actual condi-
            tions, acfm
          = velocity of gas stream in stack, ft/min
                Example Case
 In the example case, since no specific data for
 the stack are available, use the above assump-
 tions to cost the stack. The stack diameter  is
 calculated according to item (3) above. The actu-
 al gas flow rate exiting the heat exchanger (flue
 gas  flow  rate) equals approximately 40,000
 acfm. Therefore, the stack diameter is calculated
 as follows:

 Ds,ack = 0.2141  (40,000)* = 43 in

 With the stack diameter known, use the appro-
 priate  curve in Figure 5-18  (use the closest
 curve:  42  inches) to estimate the stack cost as
 follows:

 $4,500 x (336.2/226.2) = $6,700

 (Note:  12/77 dollars escalated to reflect 6/85 dol-
 lars.)
5.2.3 Estimation of the Total Purchased
     Equipment Cost
The total purchased  equipment cost equals the
sum of the major  equipment purchased cost and
the auxiliary equipment purchased cost. (Note: The
major  equipment purchased cost curves for ther-
mal incinerators, carbon adsorbers, and venturi
scrubbers also include the cost of instrumentation
and controls; therefore, this cost must be subtract-
ed to estimate the total purchased equipment cost
for these control devices.) Calculate the total pur-
chased equipment cost for  thermal incinerators,
carbon adsorbers, and venturi scrubbers as fol-
lows:  (1) multiply  the  summation  of the  major
equipment purchased cost and the auxiliary equip-
ment purchased cost by a factor of 0.091 to obtain
the cost of instrumentation and controls (the cost
of instrumentation and costs is estimated to equal
10 percent of the total purchased equipment); and
(2)  subtract this cost  from the summation of the
major  equipment purchased cost and the auxiliary
equipment purchased cost.
               Example Case
For the example case, the total purchased equip-
ment cost is estimated as follows:

(a) Total purchased equipment cost plus cost of
   instrumentation and controls  (included in
   cost curve)—
   $145,700 + $104,400 + $7,700 + $6,700 =
   $264,500

(b) Cost of instrumentation and controls—
   $264,500x0.091 = $24,100

(c) Total purchased equipment cost—
   $264,500-$24,100 = $240,400
                                        5.2.4  Estimation of Instrumentation and Controls
                                              Plus Freight and Taxes
                                        For the majority of control equipment, instrumen-
                                        tation costs are a small part of the total purchased
                                        cost.  Instrumentation requirements for  a  control
                                        system depend  upon control  and  safety require-
                                        ments.  When no specific cost data are  available,
                                        estimate the instrumentation and controls costs at
                                        10 percent of the total purchased equipment cost.
                                        The cost of equipment freight and taxes depends
                                        upon the location of the control system and the
                                        location of the supplier. Without specific data, esti-
                                        mate the cost of freight and taxes  at 8 percent of
                                        the total purchased equipment cost.
                                                         Example Case
                                          For the example case, the cost of instrumenta-
                                          tion and  controls  and the cost of freight and
                                          taxes are  as follows:
                                          (a) Instrumentation and controls—
                                             $240,400x0.10 = $24,000

                                          (b) Freight and taxes—
                                             $240,400x0.08 = $19,200
                                        5.2.5 Estimation of Total Purchased Cost
                                        The summation of the total purchased equipment
                                        cost, the cost of instrumentation and controls, and
                                        the cost of freight and taxes equals the total pur-
                                        chased cost.
                                                         Example Case
                                          The total purchased cost for the example case is
                                          as follows:

                                          $240,400 +  $24,000 + $19,200 = $283,600
                      706

-------
5.2.6 Calculation of Total Capital Costs
The  sum of the total purchased equipment cost,
other direct costs, indirect costs, and contingency
costs represents the total capital cost. Obtain the
total capital cost for the control system by multiply-
ing the total purchased cost by the  appropriate
factor listed in Table 5-8. This factor accounts for
the other direct costs, the indirect costs, and the
contingency costs. For control systems employing
multiple control devices, use the largest applicable
factor. Retrofit applications will likely be more ex-
pensive (see footnote e, Table 5-8). Each compo-
nent that comprises the total capital cost is listed in
Table 5-8 to allow insertion of specific cost data if
they are known.
                 Example Case
 The example case consists of a control system
 using one control device: a thermal incinerator.
 Therefore, by using the factor of 1.63 for thermal
 incinerators from Table 5-8, the total capital cost
 for the example case is as follows:

 $283,600 x 1.63 = $462,300
5.3  Annualized Operating Costs
The  annualized  cost  of a  control system can be
divided into direct operating costs, indirect operat-
ing costs, and credits. In this handbook, the infla-

Table 5-8.    Capital Cost Elements and Factors" (2)
tion  effect on costs  is not considered, annualized
costs are assumed to be constant in real dollars,
and  the  total annualized cost is  estimated  on a
before-tax basis. The direct operating costs consist
of utilities, operating labor charges, maintenance
charges, and replacement parts and labor charges.
Utilities (i.e., fuel, electricity, water, steam, and ma-
terials required for the control system) are annual
costs that vary depending upon the control system
size and  operating time. They are  calculated using
gas stream characteristics and control equipment
capacity data. Operating labor costs consist of op-
erator labor and supervision, while maintenance
costs consist of maintenance labor and materials.
The direct operating costs are established by esti-
mating annual quantities of utilities consumed and
operator and  maintenance labor used and by ap-
plying unit costs to these quantities. The annual
quantities  of utilities and labor requirements are
assumed to be proportional to the annual  operat-
ing hours for the control  system.  Operating  labor
supervision and maintenance materials are taken
as percentages of the operator and maintenance
labor costs. Costs of replacement parts are estimat-
ed as applicable, and the cost of replacement labor
is assumed to equal the cost of replacement parts.
The  indirect  operating  costs include overhead
costs, property  tax,  insurance,  administration
costs, and  the capital  recovery costs. Overhead
costs are estimated as a percent of operating labor


   Control Technique
Cost Elements
DIRECT COSTS
Purchased Equipment Costb
Other Direct Costs:
Foundation and supports
Erection and handling
Electrical
Piping
Insulation
Painting
Total Direct Cost
INDIRECT COSTS
Engineering and
supervision
Construction and
field expenses
Construction fee
Start up
Performance test
Model study
Total Indirect Cost
CONTINGENCY0
TOTAL"
ESP

1.00

0.04
0.50
0.08
0.01
0.02
0.02
1.67


0.20

0.20
0.10
0.01
0.01
0.02
0.54
0.07
2.27
Venturi
Scrubbers

1.00

0.06
0.40
0.01
0.05
0.03
0.01
1.56


0.10

0.10
0.10
0.01
0.01
—
0.32
0.06
1.94
Fabric
Filters

1.00

0.04
0.50
0.08
0.01
0.07
0.02
1.72


0.10

0.20
0.10
0.01
0.01
—
0.42
0.07
2.21
Thermal &
Catalytic
Incinerators

1.00

0.08
0.14
0.04
0.02
0.01
0.01
1.30


0.10

0.05
0.10
0.02
0.01
—
0.28
0.05
1.63
Adsorbers

1.00

0.08
0.14
0.04
0.02
0.01
0.01
1.30


0.10

0.05
0.10
0.02
0.01
—
0.28
0.05
1.63
Absorbers

1.00

0.12
0.40
0.01
0.30
0.01
0.01
1.85


0.10

0.10
0.10
0.01
0.01
—
0.32
0.07
2.24
Condensers

1.00

0.08
0.14
0.08
0.02
0.10
0.01
1.43


0.10

0.05
0.10
0.02
0.01
—
0.28
0.05
1.76
"As fractions of total purchased equipment cost. They must be applied to the total purchased equipment cost.
bTotal of purchased costs of major equipment and auxiliary equipment and others, which include instrumentation and controls at
 10%, taxes and freight at 8% of the equipment purchase cost.
°Contingency costs are estimated to equal 3% of the total direct and indirect costs.
dFor retrofit applications, multiply the total by 1.25.
                                                                         707

-------
and supervision  costs plus maintenance labor
costs.  Property tax, insurance, and administrative
costs are estimated as a percent of the total capital
cost. The capital  recovery cost is estimated as the
product  of  the capital recovery factor times the
total capital cost. The factor for  capital recovery
costs (the total of annual depreciation and interest
on capital) is determined from the expected life of
the control device and the interest rate at which the
capital is borrowed.  The expected  life of a given
control device depends on the type of control appli-
cation, maintenance  service, and operating duty.
For costing  purposes, preestablished expected life
values are used.

Some control techniques recover the HAP's from a
given emission stream as a salable product. There-
fore, any cost credits  associated with the recovered
material  must  be deducted from  the total annua-
lized cost to obtain the net annualized cost for the
system. The amount, purity, and commercial value
of the recovered material determine the magnitude
of credits.

5.3.1 Direct Operating Costs
Table 5-9 presents June 1985 unit costs for utilities,
Table 5-9.    Unit Costs to Calculate Annualized Cost
Cost Elements
     Unit Costs/Factor
DIRECT OPERATING COSTS
1.  Utilities:3
   a. Natural gas
   b. Fuel oil
   c. Water
   d. Steam
   e. Electricity
   f. Solvent
2.  Operating Labor:
   a. Operator Labor
   b. Supervision
3.  Maintenance:
   a. Labor
   b. Materials6
4.  Replacement:
   a. Parts
   b. Labor
INDIRECT OPERATING COSTS
1.  Overhead
2.  Property Tax
3.  Insurance
4  Administration
5.  Capital Recovery
CREDITS
$0.00425 per ft3 (11)
$1.025 per gal (12)
$ 0.0003 per gal  (2)
$0.00504 per Ib  (2)
$ 0.059 per kWh  (13)
As applicable

$11.53 per hour (14)
15% of Operator Labor

$ 11.53 per hour (14)
100% of Maintenance Labor

As applicable (see Table 5-10)
100% of Replacement Parts

80% of 2a + 2b  + 3a + 4a
1% of Total Capital Cost
1% of Total Capital Cost
2% of Total Capital Cost
(CRFC) x Total Capital Cost
As applicable
a Refer to Tables 5-10 and 5-11 to estimate utility costs for each
 HAP control technique.
bMaintenance materials include operating supplies (e.g., lubri-
 cation, paper).
CCRF = capital recovery factor.
 For an average interest rate of 10%, the CRF for specific control
 devices are listed below.
 ESP and fabric filter: CRF = 0.117 (based on 20-year life span).
 Venturi  scrubber, thermal and catalytic incinerators, adsorber,
 absorber, and condenser: CRF = 0.163 (based on 10-year life-
 span).

                      708
operator labor, and maintenance labor as well as
cost factors for other direct  operating cost ele-
ments. The procedure used to estimate direct oper-
ating costs (including utilities, direct labor, mainte-
nance, and  replacement costs) and  indirect
operating costs (including overhead, property tax,
insurance,  administration,  and  capital  recovery
cost) was  taken directly from reference 2. These
unit costs and cost factors are applied to estimated
quantities  of utilities consumed, labor expended,
and parts   used to obtain total  direct operating
costs.

If a given   control system contains two  or  more
control devices, the direct operating costs must be
calculated  for each device and summed. The cap-
ital recovery cost for a multiple control device sys-
tem should be calculated using a weighted average
capital cost factor.

Unless specified, use 8,600 hours per year, 8 hours
per shift, and 24 hours per day, as necessary, to
estimate the annual costs for utilities consumed,
operator labor, and maintenance labor.

5.3.1.1 Determine Utility Requirements
The utility requirements for a control system are
obtained from each component's design calcula-
tions.  Use  the  costing  information in  Table 5-9,
Table  5-10, and Table 5-11 to estimate the total
utility costs. A procedure to estimate fan electricity
costs is provided below, since these costs are appli-
cable to all control techniques. The fan horsepower
requirements are calculated as follows:

Fan horsepower, HP =  °-000157 x Qfg,a x AP  (5-3)

where:
HP =  fan horsepower requirement, hp
AP =  pressure drop across the control system,
       in H2O
TI   =  fan efficiency (usually 60-70%)

Assuming a 65 percent fan efficiency and a 10 per-
cent additional capacity requirement for miscella-
neous purposes, and using the conversion  factor of
0.746 kilowatt hour per horsepower-hour, estimate
the fan electricity requirement as follows:
                                            (5-4)
FER  = 2.0 x 10-4 (Qfg,a) (AP) (HRS)

where:
FER  = fan electricity requirement, kWh
HRS = hours of operation per year
                           5.3.1.2 Determine Remaining Direct Operating
                                  Costs
                           The remaining direct operating  costs include re-
                           placement parts and labor, operating labor (i.e., the
                           summation of operator labor and supervision la-
                           bor),  and maintenance (i.e., the summation  of

-------
                  Example Case
  For the thermal incinerator example case, since
  no specific information is  available, the control
  system is assumed to operate 8,600 hours per
  year. According to the  design  calculations in
  Section 4.2, and Table 5-10, the only utility re-
  quirement for the thermal incinerator, in addi-
  tion to fan electricity, is 330 scfm of natural gas.
  Using Equation  5-4, the  electricity requirement
  for a 40,000 acfm fan at 7 in H20 pressure drop is
  as follows:

  2.0  x TO'4 (40,000 acfm)  (7 in)  (8,600  hr)  =
  481,600kWh

  Applying the unit  costs from Table 5-9 and the
  equation for fuel requirements from Table 5-11,
  the utility costs are:

  (a) Natural gas cost—
     60  (330  scfm)(8,600  hr)($0.00425/ft3)  =
     $723,700

  (b) Electricity cost—
     481,600 kWh ($0.059/kWh) = $28,400
maintenance labor and materials). Tables 5-9 and
5-10 provide the necessary information to calculate
the cost of replacement parts and labor. Table 5-12
presents available data on estimated labor require-
ments for various control devices. The  labor  re-
quirements presented as "hours per shift" must be
converted to annual requirements. Total annual op-
erator and maintenance labor costs are obtained by
multiplying the  estimated annual labor  require-
ments with the applicable unit costs from Table 5-9.
These costs must be determined  for each control
device  in the  control system. Operating  labor su-
pervision and maintenance materials are estimated
as a percentage of operator labor and maintenance
labor, respectively. Again, these costs must be de-
termined for each control device if a  multiple con-
trol device system is used.
Table 5-10.   Utility/Replacement Operating Costs for HAP
            Control Techniques"
HAP Control Device
                     Utilities/Replacement Parts
Thermal Incinerator

Catalytic Incinerator


Carbon Adsorber Systems



Absorber Systems

Condenser System

Fabric Filter Systems

Electrostatic Precipitators
Venturi Scrubbers
                     Natural gas or fuel oilb
                     Electricity (fan)
                     Catalyst costc (Vcat)
                     Natural gas or fuel oilb
                     Electricity (fan)
                     Carbond (Creq)
                     Steam b
                     Cooling waterb
                     Electricity (fan)
                     Absorbentb (water or solvent)
                     Electricity (fan)
                     Refrigerant6 (Ref)
                     Electricity (fan)
                     Bags' (Atc)
                     Electricityb (fan + control device)
                     Electricity15 (fan + control device)
                     Water"
                     Electricity (fan)
a Refer to Table 5-9 for utility unit costs, Tables 5-3 and 5-5 for
 replacement part unit costs, and Table 5-2 for FE cost indices.
b See Table 5-11.
cAnnualized  replacement catalyst costs are calculated as fol-
 lows:
Annualized cost =  Vcat
                        x $/ft3
                               (Current FE/Base FE)
                     o years

dAnnualized replacement carbon costs are calculated as fol-
 lows:
  Annualized cost =
Crea (Ib) x $/lb
   5 years
                            (Current FE/Base FE)
e Refrigerant replacement is due to system leaks, however, the
 loss rate of refrigerant is very low and varies for every unit.
 Therefore, assume that the cost of refrigerant replacement is
 negligible.
'Annualized replacement bag costs are calculated as follows:
Annualized cost = A'c
                              
-------
Table 5-11.    Additional Utility Requirements (2)
Fuel Requirement for Incinerators, ft3
(Note: The design  sections for thermal and catalytic inciner-
      ators are developed under the assumption that natural
      gas is used as the supplementary fuel. Fuel oil could be
      used, however, the use of natural gas is normal industry
      practice. If fuel oil is used, the equation below can be
      used by replacing Qf with the fuel oil flow rate in units of
      gallons per  minute. The  product of the  equation then
      equals gallons of fuel oil.)
Fuel Requirement = 60 (Qf) x HRS
where:
Qf  = supplementary fuel required, scfm
MRS = annual operating hours, hr
      (Note: Use 8,600 hours unless otherwise specified.)

Steam Requirement for Carbon Adsorber, Ib
(Note: Assume 4 Ib of steam required for each Ib of recovered
      product.)
Steam Requirement = 4 (Qrec) x MRS
where:
Qrec - quantity of HAP recovered, Ib/hr
MRS = annual operating hours, hr
      (Note: Use 8,600 hours unless otherwise specified.)

Cooling Water Requirement for Carbon Adsorber, gal
(Note: Assume 12  gal of cooling water required per 100 Ibs
      steam.)
Water Requirement = 0.48 (Qrec) x HRS
where:
Qrec = quantity of HAP recovered, Ib/hr
HRS = annual operating hours, hr
      (Note: Use 8,600 hours unless otherwise specified.)
                                   Absorbent Requirement for Absorbers, gal
                                    (Note: Assume no recycle of absorbing fluid [water or solvent].)
                                    Absorbent Requirement = 60 (Lgai) x HRS
                                    where:
                                    L9ai  = absorbing fluid flow rate, gal/min
                                    HRS = annual operating hours, hr
                                          (Note: Use 8,600 hours unless otherwise specified.)

                                    Water Requirement for Venturi Scrubbers, gal
                                    (Note: Assume 0.01 gal H20 are required per acf of emission
                                          stream.)
                                    Water Requirement = 0.6 (Qe,a) x HRS
                                    where:
                                    Qe,a = emission stream flow rate into scrubber, acfm
                                    HRS = annual operating hours, hr
                                          (Note: Use 8,600 hours unless otherwise specified.)

                                   Baghouse Electricity Requirement, kWh
                                    (Note: Assume 0.0002 kW are required per ft2 of gross cloth
                                          area.)
                                    Baghouse Electricity Requirement = 0.0002 (Atc) x  HRS
                                    where:
                                    Atc  = gross cloth area required, ft2
                                    HRS = annual operating hours, hr
                                          (Note: Use 8,600 unless otherwise specified.)

                                   ESP Electricity Requirement, kWh
                                    (Note: Assume 0.0015 kW are required per ft2 of  collection
                                          area.)
                                    ESP Electricity Requirement = 0.0015 (Ap) x HRS
                                    where:
                                    Ap  = collection plate area, ft2
                                    HRS = annual operating hours, hr
                                          (Note: Use 8,600 unless otherwise specified.)
Table 5-12.    Estimated Labor Hours per Shift and Average
             Equipment Life (2)
CRF-
            + i)n]/[(1 + i)n-1]
                                                                                  (5-5)
                      Labor Requirements
                          (hr/shift)
                                          Average
     Control Device
Operator Maintenance Equipment Life
 Labor     Labor        (yr)
Electrostatic Precipitator 0.5-2
Fabric Filter 2-4
Venturi Scrubber 2-8
Incinerator 0.5
Adsorber 0.5
Absorber 0.5
Condenser 0.5
0.5-1
1 -2
1 -2
0.5
0.5
0.5
0.5
20
20
10
10
10
10
10
5.3.2 Indirect Operating Costs
The  indirect operating  costs  include overhead
costs, property tax, insurance, administration, and
the capital  recovery  cost.  Estimate the overhead
costs as 80 percent  of  the direct labor cost (the
summation of operating labor and supervision  la-
bor costs)  and the  maintenance labor cost. The
property tax estimate is calculated as 1 percent of
the total capital cost, insurance as 1 percent of the
total capital cost, and  administration as 2 percent of
the total capital cost.  Estimate the capital recovery
cost  portion of the fixed capital charges by multi-
plying  the total capital cost by  a capital recovery
factor. The capital  recovery factor (CRF) is calculat-
ed as follows:
where:
i  = interest rate on borrowed capital, decimal
n = control device life, years

For the purpose of this handbook, an interest rate
of 10 percent is used. Table 5-12 contains data on
expected control device life (n). Calculated capital
recovery factors  at  10  percent  interest  rate  are
0.163, and 0.117 for 10- and 20-year control device
lifetimes, respectively. If more than one control de-
vice is used by the control system, use a weighted
average capital recovery factor. A weighted aver-
age capital recovery factor (CRFw) is determined as
follows:
      = CRF,  [Pd/IPd  + PC2)]
        CRF2  [PC2/(Pd  + PC2)1
                                                                                   (5-6)
                                   where:
                                   CRFi = the capital recovery factor for control
                                           device #1
                                   CRF2 = the capital recovery factor for control
                                           device #2
                                   Pd  = the purchased equipment cost for control
                                           device #1
                                   PC2  = the purchased equipment cost for control
                                           device #2
                         770

-------
                 Example Case
 As estimated in Section 5.2.5, the total capital
 cost for the thermal incinerator control system
 example is $462,300. Section 5.3.1.2 estimated
 that the direct and  maintenance labor cost is
 $13,300 (i.e., $6,200 + $900  + $6,200). There-
 fore, the indirect operating costs are estimated
 as follows:

 (a) Overhead costs—
     $13,300x0.80  = $10,600

 (b) Property tax—
     $462,300 x 0.01 = $4,600

 (c) Insurance—
     $462,300 x 0.01  = $4,600

 (d) Administration charges—
     $462,300 x 0.02 = $9,200

 (e) Capital recovery—
     The capital  recovery factor is calculated us-
     ing Equation 5-5: obtain the expected equip-
     ment lifetime for an incinerator ffrom Table
     5-12 (10 years), and assume an interest rate
     of 10 percent.

     [0.1  (1 + 0.1)10]/[(1 +  0.1)10-1] = 0.163

 The capital  recovery cost  is then estimated as
 follows:

 $462,300x0.163 = $75,400

 (f) Total indirect operating costs—
     $10,600  +  $4,600 +  $4,600  +  $9,200  +
     $75,400 = $104,400
5.3.3 Credits
Credits for recovery of a salable product or energy
must be included in determining the net annualized
cost of the control system. The design calculations
for the specific control devices include the quantity
of recovered product. This information, along with
product cost data  from the  inquirer/applicant, is
used to calculate the credits on an annual basis.
                 Example Case
 For the example case, there are no  recovered
 products since a thermal incinerator destroys
 the organic vapors contained  in the emission
 stream.
                 Example Case
 Total direct and indirect operating costs for the
 example case are $771,600 and $104,400, re-
 spectively. There are no recovery credits. Thus,
 the net annualized cost of the example HAP con-
 trol system is as follows:

 $771,600 + $104,400 - $0 = $876,000
5.3.4 Net Annualized Costs
The direct and indirect operating costs less credits
received equal the net annualized cost of the HAP
control system.
5.4 References
 1.  U.S. EPA. Identification, Assessment, and Con-
    trol of Fugitive Paniculate Emissions.  Draft fi-
    nal report. EPA Contract No. 68-02-3922. April
    10, 1985.

 2.  U.S. EPA. Capital and Operating Costs of Se-
    lected Air Pollution  Control  Systems. EPA-
    450/5-80-002. December 1978.

 3.  U.S.  EPA. Organic  Chemical Manufacturing
    Volume 4: Combustion Control Devices. EPA-
    450/3-80-026. December 1980.

 4.  Vatavuk, W.M. and Neveril, R.B. "Estimate the
    Size and Cost of Incinerators." Chemical Engi-
    neering. July 12, 1982.

 5.  Vatavuk, W.M. and Neveril, R.B. "Costs of Car-
    bon Adsorbers."  Chemical Engineering. Janu-
    ary 23, 1983.

 6.  Vatavuk, W.M. and Neveril, R.B. "Costs of Gas
    Adsorbers."  Chemical  Engineering. October
    12, 1982.

 7.  U.S.  EPA. Organic Chemical  Manufacturing.
    Volume 5: Adsorption, Condensation, and Ab-
    sorption Devices. EPA-450/3-80-077. December
    1980.

 8.  Chemical  Marketing Reporter. Schnell Publish-
    ing Company. New York, NY. Volume 228. Sep-
    tember 16, 1985.

 9.  Telecon. Charlotte  Clark (PES) to  David
    Chalmers (Calgon Corp.). June 12,1985.

10.  Telecon. Gunseli Sagun Shareef (Radian Corp.)
    to Roy Uhlman  (Engelhart  Specialty Chemi-
    cals). September 30, 1985.

11.  Phillip Gennarelli, Manager, American Gas As-
    sociation (AGA). Monthly Gas Utility Statistical
    Report. AGA, 1515 Wilson Blvd., Arlington, Va.
    22209. July 1985.

12.  Bureau of Labor  Statistics. August 1985. U.S.
    consumer price for fuel oil.

13.  Monthly Energy Review. October 1982 electric-
    ity costs. DOE/EIA-003583/01. January 1983.

14.  Survey of Current Business. November 1982.
    Primary Metal Industry Costs. December 1982.
                                                                      777

-------
                                                        Appendix A. 1
                                Potential HAP's for Solvent Usage Operations

Aliphatic
Hydrocarbons

Aromatic
Hydrocarbons



Halogenated12
Hydrocarbons
Alcohols
Glycols
Ethers
Epoxides
Phenols


1 f
OI 13 0) m
,E  £ | 1 8 1 § = |
QS<0?2~-t
sS.y^llaSS
l^att^i^ll
COQO$tOtO(/)l0(/3
Specific Compounds
Cyclohexane •
Generic Compounds
Stoddard solvent •
Alicyclics • •
Specific Compounds
Benzene « •

Napthalene"
Generic Compounds

Specific Compounds
Chloromethane11
Methylene chloride • •
Chloroform"
Carbon tetrachloride • • •
1,1-dichloroethane
Trichloroethylene • •
1,1,1-trichloroethane
Tetrachloroethylene •
Trichlorotrifluoroethane • •
Chlorobenzene
o,p-Dichlorobenzene
Generic Compounds
Halogenated solvents • •
Spec/fie Compounds
Methanol • • •
Ethylene glycol • • • •
Propylene oxide
Cresols
Phenol •
Generic Compounds
Glycols •
Ethers • • • •
Phenols •
Epoxides
01 .±
CO U) »-
"5 ? 5 ^ 3
-S ~ .9 £ o
H 1 1 1 1 1 1 1
* 1= 1 ? 5 i * 3 .?
Or^ fc-O^-Crf"Oo) ^
'S ? 8 v « -c j; o c 
-------
                                                                                     I   §
                                                           V  IB   !
                                                            S;p3
                                                            ccu-
                                                                                                    o
                                                                                                    o
                                                                                                           «  •=
                                                                                                           .5
                                         •5
                                         O)
Ketones          Specific Compounds
Aldehydes          Formaldehyde
                   Acetaldehyde11
                   Furfural
                   Acetone
                   Acrolein (propenal)11
                   Methyl ethyl ketone
                   Methyl isobutyl ketone
                   Cyclohexanone
                 Generic Compounds
                   Aldehydes
                   Ketones
Esters
Amides
                 Specific Compounds
                   Ethyl acetate
                 Generic Compounds
                   Esters
                   Amides
                   Nitrosamines
Particulates
                 Specific Compounds
                   Cadmium
                   Chromium
                   Lead
                   Zinc
Acids
Nitriles
                 Specific Compounds
                   Nitrobenzene11
                 Generic Compounds
                   Organic acids
                   Nitriles
                   Nitrocompounds
Heterocyclic
  Compounds
                 Specific Compounds
                   Tetrahydrofuran
                   Furfural
                 Generic Compounds
                   Pyrrolidones
Miscellaneous
  Trade Solvents
                   So Cal I  + II
                   Solvesso 100 + 150
                   Panasolve
                   Hi Sol 100
                   TennecoT-125
                          774

-------
                                        Appendix B. 1
                           Gas Stream Parameters Calculations
At many plants, it is common that one pollution
control system serves several emission sources. In
such situations, the combined emission stream pa-
rameters must be calculated from mass and heat
balances. Procedures for calculating the combined
emission stream  and single emission stream pa-
rameters listed below are provided in this appen-
dix.
B.1.1 Flow Rate and Temperature
B.1.2 Moisture Content, S03 Content, and  Dew
     Point
B.1.3 Particulate Matter Loading
B.1.4 Heat Content

B.1.1.  Emission Stream Flow Rate and
       Temperature Calculations
Only gas volumes at  standard conditions  (70°F,
1 atm) can be added together. Thus, volumes of all
gas streams must first be converted to volumes at
standard conditions. This calculation is shown be-
low. (Note: It is assumed that the emission streams
are approximately  at  atmospheric conditions;
therefore, pressure corrections are not necessary.)
Qe1,a X
         530
      460 + Tel
= 0,
                   •el
where:
Qei,a = flow rate of gas stream #1 at actual condi-
       tions (acfm)
Tei  = temperature of gas stream #1 (°F)
Qe1  = flow rate of gas stream #1 at standard condi-
     tions (scfm)

This calculation is repeated for each emission
stream which, when combined, will be served by
the  control system. The total gas stream volumetric
flow rate at standard conditions (Qe) is calculated
by adding all gas streams, as follows:

Qe1  + Qe2 + ... = Qe

where:
Qe = flow rate of combined gas stream (scfm)

The temperature of the combined gas stream (Te)
must be calculated to convert this combined volu-
metric flow rate at standard conditions (Qe) to actu-
al conditions (Qe,a)-

The temperature of the combined gas stream (Te)
is determined by first calculating the enthalpy (sen-
                                sible heat content) of each individual stream. The
                                calculation procedures are shown below.

                                n    0.018 Btu    ,T     7n. _ H
                                 e1    ft3_°F   X ( I e1  - /U) - Hs1

                                where:
                                Tei = temperature of gas stream  #1 (°F)
                                Hs1 = sensible heat content of gas stream #1 (Btu/
                                      min)

                                This calculation is repeated for each emission
                                stream. The total sensible heat is calculated as fol-
                                lows:
H.i +
                                           ... = He
                                where:
                                Hs =  sensible heat of combined gas stream (Btu/
                                      min)

                                The combined gas stream temperature (Te) is cal-
                                culated as follows:
                                u  v
     ft3 - °F
     ll    r
                                                 1
      _      _ _
 s  0.018 Btu    Qe    e

where:
Te = temperature of combined gas stream (°F)

The actual combined gas  stream volumetric flow
rate at actual conditions (Qe,a) is then determined
as follows:
                                Q
                                  e
                                    460
                                       530
                                                *e,a
                                where:
                                Qe,a = flow rate of combined gas stream at actual
                                      conditions (acfm)
                                B.1.2  Moisture Content, SO3 Content,
                                       and Dew Point Calculations
                                Moisture content is typically reported as a volume
                                percent. The calculation procedures require that
                                the volume percent moisture content of each
                                stream be converted to a Ib-mole basis, added to-
                                gether, and then divided by the total combined gas
                                stream volumetric flow rate (Qe) to obtain the mois-
                                ture  content of the combined gas  stream. The
                                moisture content  is calculated below both on a
                                volume percent and mass percent basis. The mass
                                basis is to  allow for the dew point calculation.
                                             775

-------
The moisture content is converted from a vol %
basis to a Ib-mole basis as follows:
M
        1
  e1
      100%
Qe1 X
Ib-mole
414scf
                     ~ M
                                el.lm
where:
Me1 = moisture content of gas stream #1 (% vol)
Mei,im = moisture content of gas stream  #1 (Ib-
         mole/min)
This calculation  is repeated for each emission
stream to be combined. The moisture content of
the combined gas stream on a volume percent ba-
sis (Me)  is calculated by adding, as follows:

Me1,lm + Me2,|m + .- =  Me,|m
where
Me,im = moisture content of combined gas stream
        (Ib-mole/min)
Me = moisture content of combined gas stream (%
      vol)
The moisture content of the combined stream must
be  reported on a mass basis (Me,m) to determine
the dew point. This is calculated as follows:
M
  e,lm
 18 Ib     ..
Ib-mole ~   e-m
where:
Mem =  moisture content of combined gas stream
        (Ib/min)

The amount of dry air in the combined gas stream
(DAe) is calculated as follows:
         Q   Ib-mole     29 Ib   _ DA
         UeX 414scf    Ib-mole      e

         where:
         DAe = dry air  content of combined gas  stream
                 (Ib/min)

         Calculate the psychrometric ratio as follows:
         Me,m/(DAe - Me/m) =  psychrometric ratio (Ib of wa-
                             ter/lb dry air)

         Knowing the psychrometric ratio and the gas
         stream temperature, the dew point temperature is
         selected from Table B.1-1.

         The presence of sulfur trioxide (S03) in the gas
         stream increases the dew point of the stream. If the
         S03 component is ignored during  the dew  point
         determination, condensation may occur when not
         expected.  In addition to the problems associated
         with the entrainment of liquid droplets  in the gas
         stream, the S03 will combine with the water drop-
         lets to form su If uric acid, which causes severe cor-
         rosion on metal  surfaces and deterioration of many
         fabrics used in baghouses. Therefore, the determi-
         nation of the stream dew point must consider the
         presence of S03. With information on the S03 con-
         tent (ppm vol) and the moisture content  (% vol) of
         the gas  stream, the  "acid" dew point temperature
         can be determined from Figure B.1-1. Figure B.1-1
         provides dew points for two  moisture levels, how-
         ever, dew points can be estimated for other mois-
         ture values.

         The SO3 content of  a combined gas stream is cal-
         culated by first converting the S03 concentration of
Table B.1-1   Dew Point Temperatures
    Psychrometric
       Ratio
                                          Gas Stream Temperatures (°F)
                    70
                   80
                    90
                    100
                                       120
             140
       160
                     180
                                                             200
                                                             220
                                                240
       0.000
       0.005
       0.010
       0.015
       0.020
       0.025
       0.030
       0.035
       0.040
       0.045
       0.050
       0.055
       0.060
       0.065
       0.070
       0.075
       0.080
       0.085
       0.090
       0.095
             0
            54
            62
            68
             0
             58
             65
             72
             77
              0
             61
             68
             75
             80
             85
             89
 0
65
71
77
82
87
91
95
98
                                           Dew Point Temperatures i
  0
 70
 77
 82
 87
 91
 95
 98
101
104
107
109
111
114
116
118
119
  0
 76
 82
 86
 91
 94
 98
101
104
107
109
112
114
116
118
120
122
123
124
128
  0
 81
 86
 90
 94
 98
100
104
107
109
112
114
116
118
120
122
123
125
130
140
  0
 86
 90
 94
 97
101
104
107
109
112
114
116
118
120
122
124
125
130
140
165
  0
 89
 94
 97
100
103
107
109
111
114
116
118
120
122
123
125
130
143
162
180
  0
 93
 97
100
103
106
109
110
114
116
118
120
122
124
125
130
140
168
180
205
  0
 96
100
103
106
109
111
114
116
118
120
122
124
125
130
150
170
182
205
225
                     776

-------
Figure B.1-1. "Acid" dew points in stack gases.
    10
                        160

                   Temperature, °C
                                           200
each individual stream to a Ib-mole (Im) basis. The
S03 content is calculated as follows:

S     1     Q  x  Ib-mole _ g


where:
Sei = S03 content of gas stream #1  (ppm vol)
Sei,im = S03  content of gas stream #1 (Ib-mole/
        min)

This  is  repeated for each separate gas  stream.
These are then added to obtain the  total S03 con-
tent  of the combined gas stream  to the control
device as follows:
5e1,lm
Se,lm X
414scf    	
Ib-mole  *  QP
             = ... = S,

                 106
               e,lm
where:
Se,im = S03 content of combined  gas stream (Ib-
       mole/min)
Se =  S03 content of combined gas stream (ppm
      vol)
                                           With information for the S03 content of the com-
                                           bined gas stream (Se) and the moisture content of
                                           the combined gas stream (Me), the acid dew point
                                           is determined from Figure B.1-1.

                                           B.1.3 Participate Matter Loading
                                           Particulate matter concentrations  usually are re-
                                           ported in grains per acf. The procedures below may
                                           be used to determine  the particulate loading to a
                                           control device (in Ibs/hr) when gas streams  are
                                           combined.
                                                 Wei,B X Qel,a
                                                       v 60 min
                                                       A   f~     A
                                                           hr
                                                                           Ib
                                                                        7,000 gr
                                = WP
where:
Wel/g  =  particulate  loading  for gas  stream #1
         (gr/acf)
We1/i = particulate loading for gas stream #1 (Ib/hr)

This is repeated for each gas stream and the results
are added to obtain the particulate loading for the
combined gas stream.

              ... = We,,
                                           W
                                                   ei.|
                                                        W
                                                          e2.,
                                                 where:
                                                 We| = particulate loading for combined gas stream
                                                       (Ib/hr)

                                                 The particulate loading of the combined gas stream
                                                 can be converted to a concentration as follows:
                                                 We,,X
                                                        7,000 gr  „   1 hr
                                                                        1
                                                           Ib
                                                             60 min  A  Q,
                                                                           •= W,
                                                                              •e,a
                                                                                     e,g
                                           where:
                                           W
                                             e,g
      = particulate  loading for  combined gas
        stream (gr/acf)
B.I. 4  Heat Content Calculation
The heat content of gas stream #1 (hei) can be
determined  from the heat of combustion of its
components using the following equation:

            n
he1 =  (0.01)  2  ye1il  x he1,i
            i = 1

where:
he1 =  heat content in gas stream #1 (Btu/scf)
yeu = volume  percent of component "i" in gas
       stream #1 (% vol)
he1 , =  heat of combustion of component "i" in gas
       stream #1 : see Table B.1-2 (Btu/scf)
n = number of components in gas stream #1

The heat content of a combined emission stream
can be determined  from the  heat content of the
individual emission streams as follows:
He = (0.01)
                                                       m
                                                           yej x he
                                                                     777

-------
where:
hfi  -
Yej  =

hei  =
m =
                                n-lt
                                    in combined
 combined emission stream heat  content
 (Btu/scf)
volume percent of stream
gas stream (% vol)
heat content of stream  "j" in combined gas
stream: see previous discussion (Btu/scf)
number of individual gas streams  in com-
bined gas stream
Table B. 1 -2.   Heats of Combustion and Lower Exposive Limit
            (LEL) Data for Selected Compounds*
Compound
Methane
Ethane
Propane
n-Butane
Isobutane
n-Pentane
Isopentane
Neopentane
n-Hexane
Ethylene
Propylene
n-Butene
1-Pentene
Benzene
Toluene
Xylene
Acetylene
Naphthalene
Methyl alcohol
Ethyl alcohol
Ammonia
Hydrogen sulfide
LEL
(ppmv)
50,000
30,000
21,000
16,000
18,000
15,000
14,000
14,000
11,000
27,000
20,000
16,000
15,000
13,000
12,000
11,000
25,000
9,000
60,000
33,000
160,000
40,000
Net Heat of
Combustion"'0
(Btu/scf)
892
1,588
2,274
2,956
2,947
3,640
3,631
3,616
4,324
1,472
2,114
2,825
3,511
3,527
4,196
1,877
1,397
5,537
751
1,419
356
583
"Sources: Steam/Its Generation and Use, The Babcock & Wilcox
        Company. New York, NY. 1975.
        Fire Hazard Properties of Flammable Liquids, Gases,
        Volatile Solids -1977. National Fire Protection Associ-
        ation. Boston, MA. 1977.
lower heat of combustion.
'Based on 70°F and 1 atm.
                Example Case
Calculate the heat content of an emission stream
from a paper coating operation  (gas stream #1)
with the following composition data: methane
(44 ppmv), toluene (73 ppmv), and  others  (4
ppmv). Let subscripts "1" and "2" denote meth-
ane and toluene, respectively.
he1 = (0.01) (ye1(1 x he1<1 +  yei,2 x hei,2)

Convert  the concentrations to  volume  percent
basis:
Methane: ye1i1  =  0.0048  (assume "others" is
                          equivalent  to meth-
                          ane)
Toluene: ye1,2 = 0.0073

From Table B.1-2:
Methane: he1/1 = 892 Btu/scf
Toluene: he12 = 4,196 Btu/scf

Substituting these  values in the above equation
yields:
he1 = 0.35 Btu/scf
                                                    Table B.1-3   Properties of Selected Organic Compounds*
Compound
Acetone
Benzene
n-Butyl acetate
n-Butyl alcohol
Carbon tetrachloride
Chloroform
Cyclohexane
Ethyl acetate
Ethyl alcohol
Heptane
Hexane
Isobutyl alcohol
Isopropyl acetate
Isopropyl alcohol
Methyl acetate
Methyl alcohol
Methylene chloride
Methyl ethyl ketone
Methyl isobutyl ketone
Perchloroethylene
Toluene
Trichlorethylene
Trichlorotrifluoroethane
Xylene
Molecular
Weight
(Ib/lb-mole)
58
78
116
74
154
119
54
88
46
100
86
74
103
60
74
32
85
72
100
166
92
131
187
106
Boiling
Point
(°F)
133
176
257
243
170
142
176
171
173
209
156
225
191
181
135
148
104
175
244
250
231
189
118
281-292
                                                    "Source: Chemical Engineer's Handbook. Perry, R.H. and Chil-
                                                            ton, C.H. (eds). Fifth Edition. McGraw-Hill Book Com-
                                                            pany. New York, NY. 1973.
                       778

-------
                                         Appendix B.2
                           Dilution Air Requirements Calculations
B.2.1  Dilution Air Calculations
The quantity of dilution air (Qd) needed to decrease
the heat content of the  emission stream to hd is
given by the following equation:
   = [(IVhd)-1]Qe
(1)
where:
Qd = dilution air flow rate, scfm
he = emission stream heat content before dilution,
     Btu/scf
hd = emission stream heat content after dilution,
     Btu/scf
Qe  = emission stream flow rate before dilution,
      scfm

The concentrations of the various components and
flow rate of the emission stream have to be adjust-
ed after dilution as follows:

02,d = 02 (hd/he) + 21 [1 - (hd/he)]             (2)

Me,d =  Me (hd/he) + 2 [1 - (hd/he)]             (3)
Qe,d = Qe (he/ hd)                            (4)
where:
O2d = oxygen content of diluted emission stream,
      vol %
Me_d =  moisture content of  diluted emission
        stream, vol %
Qe/d  = flow rate of the diluted emission stream,
       scfm

The factor 21  in Equation 2 denotes the volumetric
percentage of oxygen in air  and the factor 2 in
Equation 3 is the volumetric  percentage of mois-
ture in air at 70°F and 80 percent humidity.

After dilution, the HAP emission stream character-
istics are redesignated as follows:
                                                  02
                                                  Me

                                                  he
                                                  Qe
            02,d  =
            Me,d  =
            hd   =
            Qe,d  =
                      . Btu/scf
                      .scfm
                                                  Appendix C.2 is a worksheet for calculating dilution
                                                  air requirements.
                                                                      779

-------
        Appendix C. 1
HAP Emission Stream Data Form
                               121

-------




















/I DATA FORM*
«=
UJ
CC
t,

o
t/5
c/>
^
UJ
CL

X































mt Contact
IU




















>
C
re
c
f
L.
^
C




















0
•z.
01
c
o
x:
c
a
a
r~



















^*<
\
a
(7
•-._
K
: c
1 C
i- '£
: n
^
) f




















t
CO
:^j
c
O
CJ
>
CJ
L C
03
C




















i
> ">
; ±
) C.
•• -*~
5
j
3
>
\
^
J







_a

'>
a:
nr
^^
k.
03
"a
c
D
w


a
t/
1
a
E
. LU
M-
o
^ o
^
^m



















-<— •
C
K
. a
< 0
. 4-
) cy
•» ^^












































L
i
h.
)
i
1
!































c
c
'4^
re
c
J4^
^"
C
a
"C
4-
C
re
E
—
k.
a
.c
£
z
E
re
a
4-
cr
c
c
'a
a
£
LL

<











C









X















re




!
: a
: ^
- c
It
: C
! S
II
J a
i§

: a











) C









I J2















03




•—
I to
) u
. i£
) c/
: w
1 -2
: U
i S
: 1
: co

J C











c









.c















CC




(/•
Q.
2
£
re
o
k.
t/-
c
c
V
u
E
LL

C











) C









X















re




E
k.
c
LL
!^
o
: o
> J!
i C
'§
I Z

I LL











) 0









2 .a















I CO




*
: *.
: co
> CM
• •
] c
) 1
J 0
) 0
. Q.

J LL











U









-Q















03


j.— i.
c\
*^_-
^
a
' «
CL
0
a
CO
^
a_
I

G











c









^:















CO




CS
r—
**—
>
1
L —
C
CO
CL
<
I

Z











u









_a















(0


cs

-•^^
a
- 2
- Q.
' 03
- >
C
0
w
•a
<
CL
<
I













0









JQ















CO

^— •
- CN
" *
i- a
a
?
CO
Lc3
a:
i
a.
<
X

^















*
#
*

-w
C
o;
c
0
O
0
'c
03
O
CD
oj

















'

«!*•
- o-
cs
^ ^.
"—
4-
c
a
•£
c
t-'
2:
mm
$
c
^

^

















. -^

*• •
4-<
a
o
U
r
O
4-^
) CC
a
I
a



















"
i
- cr
CS
) r-
—
£
-
i «
k.
i a
:f
: h

• _•















<.• .
CO

•t
• «
C
IL.
1



















*— «
CS
^
^«
v_
I a
,|
> ~£
1
! «
z

: c
















'


0
K




















.^~
C/
•>•»
1 1
"E
' 0
CC
_a
re
D .S:
) -£
) C
: <

i z










































h
> c!
If
: c
> ^-
U
L a

J >








































C/)
T3
C
> .c
4-
'1
> c
i £
1 c
!l
u
i 0
: t/












































>
!
)
t
)
)
j
)

»
»
O
4->
0
E
"o
'c
o :
0 LU
T3 =
0 TO
O
0 CD
0 =
CO (J
j- • CO
4^ T3 Q.
o 0 0
^ 0 4>2
c c w
~ co c
>• "c
J-: +3 CD
4^ 4^
C 0 CD i,'
05 0 "°
k. "O "4^ CD
1- § ?
J^ 05 OJ
c a. •- •-,
E < T3 -1
CD T= C c
0 -L 0 o
i; 0 Q- ._
co 0 0 E
T3 £ ^ §
C C^ 0 "^
15 CD ~
E £ a. ^
O i- CO .—
k. 0 0 t?
o E -Q 3
0 T3 -°
OJ Dl 3 ^
.E_0 o g
•*•* >. w Q-
E»- CD <
CD 4™* ^r
CD CO CD
0 CO T3 CO
| § ^ c |
.2 w ^ §'<« ^
co •- 0 ~ .<2 J2
•|S o 8 E § x
E CD c: F 0 ~ ^
0 Q. 42 S CO « ^3
c W <-! ^ CO •= C
^ _ co co 0 £ c
10 CD 05 CO O Ji^ O
The data presented are for
Use extra forms if addition
The numbers in parenthes<
1 = organic vapor proce
2 = inorganic vapor pro
3 = particulate process f
Organic emission stream c
* * *
* *
*

-------
                           Appendix C.2
           Calculation Sheet for Dilution Air Requirements

Dilution airflow rate:

        Qd =  [(he/hd)-1]Qe

        Qd =  _ scf m

Diluted emission stream characteristics:

        02,d = 02 ( hd/he) + 21 [1 - ( hd/he)]

        02,d = _ %

        Me,d = Me ( hd/he) + 2 [1 - ( hd/he)]

        Me/d = _ %
        Qe,d = Qe (he/ hd)

        Qe d = _ scf m

Redesignate emission stream characteristics:

        02 =  02,d = - %

        Me =  Me,d = _ %

        he = hd = _ Btu/scf

        Qe =  Qe,d = - scf m
                                                     723

-------
                                 Appendix C.3
                    Calculation Sheet for Thermal Incineration

4.2.1 Data Required
        HAP emission stream characteristics:3

        1. Maximum flow rate, Qe =	scfm

        2. Temperature, Te =	°F

        3. Heat content, he =	Btu/scf

        4. Oxygen content,6 02  =	%

        5. Moisture content, Me =	%

        6. Halogenated organics: Yes	No	
        Required destruction efficiency, DE =	%

In the case of a permit review, the following data should be supplied by the applicant:

        Thermal incinerator system variables at standard conditions (70°F, 1 atm):

         1.  Reported destruction efficiency, DEreported =	%

         2.  Temperature of emission stream entering the incinerator,

            Te =	°F (if no heat recovery);

            The =	°F (if a heat exchanger is employed)

         3.  Combustion temperature, Tc =	°F

         4.  Residence time, tr =	sec

         5.  Maximum emission stream flow rate, Qe =	scfm

         6.  Excess air, ex =	%
alf dilution air is added to the emission stream upon exit from the process, the data
  required are the resulting characteristics after dilution.

bThe oxygen content depends on the oxygen content of the organic compounds (fixed
  oxygen) and the free oxygen in the emission stream. Since emission streams treated by
  thermal incineration are generally dilute VOC and air mixtures, the fixed oxygen in the
  organic compounds can be neglected.
                  724

-------
         7.  Fuel heating value , hf = _ Btu/scf (assume natural gas)

         8.  Supplementary heat requirement, Hf = _ Btu/min

         9.  Combustion chamber volume, Vc = _ ft3

        10.  Flue gas flow rate, Qfg = _ scfm

        11 .  Heat exchanger surface area (if a heat exchanger is employed),

            A = _ ft2
4.2.2 Pretreatment of the Emissions Stream: Dilution Air Requirements
Typically, dilution will not be required. However, if the emission stream heat content (he)
is greater than  13 Btu/scf with oxygen concentration greater than 16  percent, see
Appendix C.2 where a blank calculation sheet for determining dilution air requirements
is provided.

4.2.3 Thermal Incinerator System Design Variables
Based on the required destruction efficiency (DE), select appropriate values for Tc and tr
from Table 4-1.
                T =           °F
                i c  - r

                tr = _ sec

For a permit evaluation, if the applicant's values for Tc and tr are sufficient to achieve the
required DE (compare the reported  values  with the values presented in Table 4-1),
proceed with the calculations. If the applicant's values for Tc and tr are not sufficient, the
applicant's design is unacceptable. The reviewer may then use the values for Tc and tr
from Table 4-1.

                T =           °F
                ic  - r
                tr =	sec

(Note: If DE is less than 98 percent, obtain information from literature and incinerator
vendors to determine appropriate values for Tc and tr.)

4.2.4  Determination of Incinerator Operating Variables

4.2.4.1 Supplementary Heat Requirements
        1. For dilute emission streams that require no additional combustion air:

                a. Use Figure 4-2:

                   Hf = (Hf/Qe)fjgUre Qe

                                                           125

-------
          Hf =	Btu/min

          or

        b. Use Equation 4.2-1:

                          + 0.002 Me) [Cpair (Tc-Tr) - Cpair (The-Tr) -^
            f~-f             hf-1.4Cpair(Tc-Tr)

       The values for the parameters in this equation an be determined as
       follows:

       Qe, he, Me       Input data.

       hf              Assume a value of 882 Btu/scf if no other information
                       is available.

       Cpair           Assume a value of 0.0190 Btu/scf-°F if no other infor-
                       mation is available.

       Tc              Obtain value from Table 4-1 or from permit applicant.

       The             Use the following equation if the value for The is not
                       specified:
                       The = (HR/100) Tc + [1 - (HR/100)] Te

                       where HR = heat recovery in the heat exchanger (per-
                       cent). Assume a value of 50 percent for HR if no other
                       information is available.

        Tr              70°F

        Hf =           	Btu/min

        If Hf is less than 5 Btu/min, redefine Hf = 5 Btu/min.

2.  For emission streams that are not dilute and require additional combustion
   air:


        Use Figure 4-3 to obtain a conservative estimate:

        Hf = (Hf/Qe)figure Qe

        Hf =	Btu/min

          726

-------
4.2.4.2 Flue Gas Flow Rate
        1. For dilute emission streams, use Equation 4.2-2:

                Qfg = Q6 + Qf + QC

                where:
                and

                Qf =  Hf/hf

                Qfg =	scf m

        2.  For emission streams that require additional combustion air, use the follow-
           ing equation to calculate Qc:

                Qc  = [(0.01 He + 9.4 Qf) (1 +  0.01 ex) - 0.0476 02 Qe]

                He  = Qehe

                Assume ex = 18 percent if no other information is available.

                Qc  =	scf m

                Then use Equation 4.2-2 to calculate Qfg:

                Qfg =	scf m

4.2.5 Combustion Chamber Volume
                a. Use Equation 4.2-4 to Convert  Qfg (standard conditions) to Qfg/a
                  (actual conditions):

                  Qfg,a = Qfg [(Tc +  460)/530]

                  (Note: Pressure effects are negligible.)

                  Qfg/a =	acfm

                b. Use Equation 4.2-5 to calculate combustion chamber volume:

                  Vc=  [(Qfg,a/60)tr]1.05

                  Obtain value for tr from Table 4-1 or from permit applicant.

                  Vc =	ft3
                                                            127

-------
                   If Vc is less than 36 ft3 (minimum commercially available size),

                   Vr = 36 ft3
                   ' c
4.2.6 Heat Exchanger Size
        1. For dilute emission streams that do not require additional combustion air:

                a.  Use Figure 4-4:

                   A =	ft2

                b.  Use Equation 4.2-6:

                   A = [60 Qe (1 + 0.002 Me) Cpair (The - Te)]/U ATLM

                   The values for the parameters in this equation can be determined as
                   follows:

                   Qe, Cpair, The, Me   As specified for Equation 4.2-1.

                   Te                Input data.

                   U                 Use a value of 4 Btu/hr-ft2-°F unless the inquirer/
                                     applicant has provided a value.

                   Tc                As specified for Equation 4.2-1.

                   ATLM              Calculate ATLM using the following expression:

                                     ATLM = Tc - The

                                     ATLM =	°F

                   Heat exchanger surface area:

                   A =	ft2
        2. For emission streams that are not dilute and require additional combustion
           air:
                   Use Figure 4-5 :

                   A = (A/Qe)figure Qe

                   A =	ft2

                   728

-------
4.2.7 Evaluation of Permit Application
Compare the calculated values and reported values using Table 4-2. The combustion
volume (Vc) is calculated from flue gas flow rate (Qfg) and Qfg is determined by emission
stream flow rate (Qe), supplementary fuel flow rate (Qf), and combustion air requirement
(Qc). Therefore, if there are differences between the calculated and reported values for Vc
and Qfg, these are dependent on the differences between the calculated  and reported
values for Qc and Qf.

If the calculated and reported values are different, the  differences may be  due to the
assumptions involved in the calculations. Discuss the details of the design and operation
of the system with the applicant.

If the calculated and reported values are not different, then the design and operation of
the system can be considered appropriate based on the assumptions employed in this
handbook.
Table 4-2.    Comparison of Calculated Values and Values Supplied by the Permit
            Applicant for Thermal Incineration
                                              Calculated    Reported
                                                Value       Value

           Supplementary heat requirement, Hf       ...         ...

           Supplementary fuel flow rate, Qf           ...         ...

           Flue gas flow rate, Qfg                    ...         ...

           Combustion chamber size, Vc             ...         ...

           Heat exchanger surface area, A            ...         ...
                                                          723

-------
                                  Appendix C.4
                    Calculation Sheet for Catalytic Incineration

4.3.1 Data Required
        HAP emission stream characteristics:3

        1. Maximum flow rate, Qe =	scfm

        2. Temperature, Te =	°F

        3. Heat content, he =	Btu/scf

        4. Oxygen content13, 02 =	%

        5. Moisture content, Me =	%

        Required destruction efficiency, DE	%

In the case of a permit review, the following data should be supplied by the applicant:

        Catalytic incinerator system variables at standard conditions (70°F, 1 atm):

         1. Reported destruction efficiency, DEreported =	%

         2. Temperature of emission stream entering the incinerator,

           Te =	°F (if no heat recovery),

           The =	°F (if emission stream is preheated)

         3. Temperature of flue gas leaving the catalyst bed,

           T   =	      °F
            i co   	  r
alf dilution air is added to the emission stream upon exit from the process, the data
 required are the resulting characteristics after dilution.

bThe oxygen content depends on the oxygen content of the organic compounds (fixed
 oxygen) and the free oxygen in the emission stream. Since emission streams treated by
 catalytic incineration are generally dilute VOC and air mixtures, the fixed oxygen in the
 organic compounds can be neglected.
                  730

-------
         4. Temperature of combined gas stream (emission stream + supplementary
           fuel combustion products) entering the catalyst bed,3
         5. Space velocity, SV =	hr"1

         6. Supplementary heat requirement, Hf =	Btu/min

         7. Flow rate of combined gas stream entering the catalyst bed,

           Qcom =	scf m

         8. Combustion air flow rate, Qc =	scfm

         9. Excess air, ex =	%

        10. Catalyst bed requirement, Vbed =	ft3

        11. Fuel heating value, hf =	Btu/scf

        12. Heat exchanger surface area (if a heat exchanger is employed),

           A =	ft2

4.3.2 Pretreatment of the Emission Stream: Dilution Air Requirements
For emission streams treated by catalytic incineration, dilution air typically will not be
required. However, if the emission stream heat content is greater than 10 Btu/scf for air
+ VOC mixtures or if the emission stream heat content is greater than 15 Btu/scf for inert
+ VOC mixures, dilution air is necessary. For emission streams that cannot be character-
ized as air + VOC or inert + VOC mixtures, assume that dilution air will be required if the
heat content is greater than 12 Btu/scf. In such cases, refer to Appendix C.2 where a blank
calculation sheet for determining dilution air requirements is provided.

4.3.3 Catalytic Incinerator System Design Variables
Based on the required destruction efficiency (DE), specify the appropriate ranges for Tci
and Tco and select the value for SV from Table 4-3.

                Tci (minimum) = 600°F

                Tco (minimum) = 1,000°F

                Tco (maximum) = 1,200°F

                SV =	hr1
alf no supplementary fuel is used, the value for this variable will be the same as that for
 the emission stream.

                                                           737

-------
In a permit review, determine if the reported values for Tci, Tco, and SV are appropriate to
achieve the required destruction efficiency. Compare the applicant's values with the
values in Table 4-3 and check if:

                Tci (applicant) > 600°F and 1,200°F > Tco (applicant) > 1,000°F

                and

                SV (applicant) < SV (Table 4-3)

If the reported values are appropriate,  proceed with the  calculations. Otherwise, the
applicant's design is considered unacceptable. The reviewer may then wish to use the
values in Table 4-3.

4.3.4 Determination of Incinerator Operating Parameters

4.3.4.1 Supplementary Heat Requirements
        1.  For dilute emission streams that require no additional combustion air:

                a.  Use Equation 4.3-1 to determine if Tci = 600°F from  Table 4-3 is
                   sufficient to ensure an adequate overall reaction rate without dam-
                   aging  the catalyst, i.e., check if Tco falls in the interval  1,000° -
                   1,200°F:

                   Tco = 600 + 50 he

                   T  =           °F
                   i co   	 r
                   If Tco fa||s in the interval 1,000° - 1,200°F, proceed with the calcula-
                   tions. If Tco is less than 1,000°F,  assume Tco is equal to 1,000°F and
                   use Equation 4.3-2 to determine an  appropriate value for Tci; and
                   then proceed with the calculations:

                   Tci  = 1,000-50he

                   T .  - _   °p
                   i c,    -  r
                   (Note: If Tco is greater than  1,200°F, decline in catalyst activity may
                   occur due to exposure to high temperatures.)

                b.  Use Figure 4-7 to determine supplementary heat requirements:

                   Hf = (Hf/Qe)figure Qe

                   Hf = _ Btu/min

                c.  Use Equation 4.3-3 to determine supplementary heat requirements:

                  732

-------
                  Hf - 1.1  hfQe(1 + 0.002 Me)
r[Cpair(Tci-Tr)-Cpair(The-Tr)]l
L   h, - 1.4 Cpair (Tci - Tr)   J
                  The values for the variables in this equation can be determined as
                  follows:

                  Qe, Me, Te       Input data.

                  hf              Assume a value of 882 Btu/scf (for natural gas) if no
                                  other information is available.

                  Cpair            Assume a value of 0.0190 Btu/scf-°F  if no other
                                  information is available.

                  Tci              Obtain  value  from  part a above  or from permit
                                  applicant.

                  The             For no heat recovery case, The = Te. For heat recov-
                                  ery case, use the following equation if the value for
                                  The is not specified:

                                  The = (HR/100)TCO + [1 - (HR/100)] Te

                                  where HR  = heat recovery in the heat exchanger
                                  (percent). Assume a value of 50 percent for HR if no
                                  other information is available.

                  Tr              70°F

                  Hf =            	Btu/min

        2.  For emission streams that are not dilute and require additional combustion
           air:

                  Use Figure 4-8 to obtain a conservative estimate:

                  Hf = (Hf/Qe)figure Qe

                  Hf =	Btu/min

4.3.4.2.  Flow Rate of Combined Gas Stream Entering the Catalyst Bed
        1.  For dilute emission streams that  require no  additional combustion air, use
           Equations 4.3-4 and -5:

                Qcom = Qe + Qf +  Qc

                  Qf = Hf/hf
                                                           733

-------
                Qf =	scf m

                Qcom =	scf m

        2.  For emission streams that require additional combustion air, use the
           following equation to calculate Qc:

                Qc = [(0.01 he Qe + 9.4 Qf) (1  + 0.01 ex) - 0.0476 02 Qe]

                Qc =	scf m

                Then use Equation 4.3-4 to calculate QCOm:

                Qcom =	scf m

4.3.4.3 Flow Rate of Flue Gas Leaving the Catalyst Bed
                a.  Use the result from the previous calculation:

                  Qfg = Qcom

                  Qfg =	scf m

                   If Qfg is less than  500 scfm, define Qfg as 500 scfm.

                b.  Use Equation 4.3-6 to calculate Qfg;a:

                   Qfg,a = Qfg KTCO + 460)7530]

                   Qfg,a =	acfm

4.3.5  Catalyst Bed Requirement
                Use Equation 4.3-7:

                Vbed = 60 Qcom/SV

                Vbed =	ft3

4.3.6  Heat Exchanger Size (for Systems with Recuperative Heat Exchange Only)
        1.  For dilute emission streams that do not require additional combustion air:

                a.  Use Figure 4-9 (line 1):

                  A = (A/Qe)figure Qe

                  A =	ft2

                b.  Use Equation 4.3-8:
                  734

-------
                A =  [60 Qe (1 + 0.002 Me) Cpair (The-Te)/UATLM]

                The values for the parameters in this equation can be determined as fol-
                lows:
                       r' The' Me, he   As specified for Equations 4.3-1 and -3.

                Te                   Input data.

                U                    Use a value of 4 Btu/hr-ft2-°F unless the inquirer/
                                     applicant has provided a value.

                Tco                  As calculated in Step 1 of 4.3.4.1 :

                                     T  - _ °P
                                     1 CO   - r
                ATLM                 Calculate ATLM using the following expression:

                                     ATLM = Tco - The

                                     ATLM =	°F

                Heat exchanger surface area:

                A =	ft2

        2. For emission streams that are not dilute and require additional combustion
           air:

                Use Figure 4-9 (line 2):

                A = (A/Qe)figure Qe

                A =	ft2
4.3.7 Evaluation of Permit Application
Compare the calculated values and the values supplied by the applicant using Table 4-4.

If the calculated values for Hf, Qc, QCOm' Vbed, ancl A differ from the applicant's values, the
differences may be due to the assumptions involved in the calculations. Discuss the
details of the design and operation of the system  with the applicant.

If the calculated and reported values are not different, then the design and operation of
the system can be considered appropriate based on the assumptions employed in this
handbook.

                                                           735

-------
Table 4-4.    Comparison of Calculated Values and Values Supplied by the Permit
             Applicant for Catalytic Incineration
                                               Calculated   Reported
                                                 Value       Value
           Supplementary heat requirement, Hf

           Supplementary fuel flow rate, Qf

           Combustion air flow rate, QfC

           Flue gas stream flow rate, CCOm

           Catalyst bed volume, Vbed

           Heat exchanger surface area,
           (if recuperative heat recovery
           is used), A
                  136

-------
                                  Appendix C.5
                           Calculation Sheet for Flares

4.4.1 Data Required
        HAP emission stream characteristics:

        1. Expected emission stream flowrate, Qe =	scfm

        2. Emission stream temperature, Te =	°F

        3. Heat content, he =	Btu/scf

        4. Mean molecular weight of emission stream MWe =	Ib/lb-mole

        Flare tip diameter, Dtip =	in

        Required destruction efficiency, DE =	%

In the case of a permit review, the following data should  be supplied by the applicant:

        Flare system design  parameters at standard conditions (70°F, 1 atm):

         1.  Flare tip diameter, Dtjp =	in

         2.  Expected emission stream flowrate, Qe =	scfm

         3.  Emission stream heat content, he =	Btu/scf

         4.  Temperature of emission stream, Te =	°F

         5.  Mean molecular weight  of emission stream, MWe =	Ib/lb-mole

         6.  Steam flowrate,  Qs =	Ib/min

         7.  Flare gas exit velocity, Ufig =	ft/sec

         8.  Supplementary fuel flow rate,3 Qf =	scfm

         9.  Supplementary fuel heat content,3 hf =	Btu/scf
aThis information is needed if the emission stream heat content is less than 300 Btu/scf.

                                                           737

-------
        10.  Temperature of flare gas,bTfig =	°F

        11.  Flare gas flowrate,b Qfig =	scfm

        12.  Flare gas heat content,6 hfig =	Btu/scf

4.4.2  Determination of Flare Operating Variables
Based on studies conducted  by EPA, relief gases having heating values less than 300
Btu/scf are not assured of achieving  98 percent destruction efficiency when they are
flared in steam- or air-assisted flares.0

In a permit review case, if  he is below 300 Btu/scf and no supplementary fuel is used,
then the  application is rejected. The reviewer  may then wish  to proceed with the
calculations below. If he is equal to or above 300 Btu/scf, then the reviewer should skip to
Section 4.4.2.3.

4.4.2.1 Supplementary Fuel Requirements
For emission streams with heat contents less than 300 Btu/scf, additional fuel is required.
Use Equation 4.4-1 to calculate natural gas requirements:

                Qf = [(300 - he) Qel/582

                Qf =	scfm

4.4.2.2 Flare Gas Flow Rate and Heat Content
                a.  Use Equation  4.4-2 to calculate the flare gas flow rate:

                   Qf|g = Q6 + Qf

                   Qfig =	scfm

                b. Determine the flare gas heat  content as follows:

                   hfig =  300 Btu/scf if Qf > 0

                   hfig =  he if Qf = 0

                   hfig =	Btu/scf

4.4.2.3 Flare Gas Exit Velocity
                a. Use Table 4-5 to calculate Umax:
blf no auxiliary fuel is added, the value for this variable will be the same as that for the
 emission stream.

c For unassisted flares, the lower limit is 200 Btu/scf.
                  738

-------
                   If 300 s hfig < 1,000, use the following equation:

                   Umax = 3.28 [io 1,000 Btu/scf, Umax  = 400 ft/sec

                b.  Use Equation 4.4-3 to calculate Ufig:

                   Uf,g = 3.06 Qfig,a/(Dtip)2

                   where Qfig,a is given by Equation 4.4-4:

                   Qflg,a =  [Qflg (Tf|g + 460)]/530

                   See Appendix B.7,  reference 8, for calculating Tfig.

                   Qfig,a =	acfm

                   Ufig =	ft/sec

                c.  Compare Ufig and Umax:

                   If Ufig < Umax/ the desired destruction efficiency level of 98 percent
                   can be achieved. (Note:  Ufig should exceed 0.03 ft/sec for flame
                   stability.) If Ufig > Umax, 98 percent destruction efficiency cannot be
                   achieved. When evaluating a permit, reject the application in such a
                   case.

4.4.2.4 Steam Requirements
                a.  Assume that the amount of steam required is 0.4 Ib steam/lb  flare
                   gas. Use Equation 4.4-5 to calculate Qs:

                   QS = 1.03XlO"3XQf|gXMWf|g

                   Qs =	Ib/min


4.4.3  Evaluation of Permit Application
Compare the calculated and reported values using Table 4-6. If the calculated values of
Qf,  Ufig, Qfig, and Qs are different from the reported values for these variables, the
differences may be due to the assumptions (e.g., heating value of fuel, ratio of steam to
flare gas, etc.) involved in the calculations. Discuss the details of the design and oper-
ation  of the system with the applicant. If the calculated and reported values are not
different, then the operation of the system can be considered appropriate based on the
assumptions employed in the handbook.
                                                             735

-------
Table 4-6.    Comparison of Calculated Values and Values Supplied by the Permit
             Applicant for Flares
                                             Calculated     Reported
                                                Value         Value


           Supplementary fuel flow rate, Qf         ...           ...

           Flare gas exit velocity, Uf|Q

           Flare gas flow rate, Qfig

           Steam flow rate, Qs                    ...           ...
                  140

-------
                                 Appendix C.6
                     Calculation Sheet for Carbon Adsorption

4.6.1 Data Required
        HAP Emission stream characteristics:

        1. Maximum flow rate, Qe =	scfm

        2. Temperature, Te =	°F

        3. Relative humidity, Rhum =	%

        4. HAP =	

        5. Maximum HAP content, HAPe =	ppmv

        Required removal efficiency, RE =	%

In the case of a permit review, the following data should be supplied by the applicant:

        Carbon adsorber (fixed-bed) system variables at standard conditions (70°F, 1 atm):

         1. Reported removal efficiency, REreported =	%

         2. HAP content, HAPe =	ppmv

         3. Emission stream flow rate, Qe =	scfm

         4. Adsorption capacity  of carbon bed,

           AC =	Ib HAP/100 Ib carbon

         5. Number of beds =	

         6. Amount of carbon required, Creq =	Ib

         7. Cycle time for adsorption, 6ad =	hr

         8. Cycle time for regeneration, 0reg =	hr

         9. Emission stream velocity through the bed, Ue =	ft/min

        10. Bed depth, Zbed =	ft

        11. Bed diameter, Dbed =	ft

        12. Steam ratio, St =	Ib steam/lb carbon

                                                         747

-------
4.6.2 Pretreatment of the Emission Stream

4.6.2.1  Cooling
If the temperature of the emission stream is significantly higher than 100°F, a heat
exchanger is needed to cool it to 100°F. Refer to Appendix  B.5, reference 8, for the
calculation procedure.

4.6.2.2 Dehumidification
If the relative humidity level is above 50 percent, a condenser is required to cool and
condense the water vapor in the emission stream. Refer to Section 4.8 for more details.

4.6.2.3 High VOC Concentrations

               HAPe = _ ppmv

If flammable vapors are present in the emission stream, VOC content will be limited to
below 25 percent of the LEL.

               LEL = _ ppmv (from Table B.1-1)

               25% of LEL = 0.25 x LEL (ppmv) = _ ppmv

The  maximum practical inlet concentration for carbon  beds is about 10,000 ppmv. If
HAPe is greater than 10,000 ppmv, carbon adsorption may not be applicable.

4.6.3  Carbon Adsorption System Design Variables
               a. Use Equation  4.6-1 to calculate the required outlet HAP concentra-
                  tion:

                  HAP0 = HAPe (1 - 0.01 RE)

                  HAP0 = _ ppmv

               b. Specify the appropriate values of 0ad, 0reg, and St from Table 4-7.

                  0ad = - hr

                  0reg = - hr

                  St = _ Ib steam/lb carbon
                  742

-------
4.6.4  Determination of Carbon Adsorber System Variables

4.6.4.1 Carbon Requirements
                a.  Use Equation 4.6-2:

                   Creq = 2 x 1.55 1(T5 N6ad Qe (HAPe - HAP0) MWHAP/AC

                   Assume N = 2

                   Obtain MWHAP  from Table B.1-2 or reference 18.  If no data  are
                   available, use a  conservative value of 5 Ib HAP/100 Ib carbon.

                   Creq = - Ib
                b.  Use Figure 4-13 to obtain (Creq/Qe):

                   Weq = ^req'^e'figure ^e

                   Creq = - Ib
4.6.4.2 Carbon Adsorber Size
                a.  Use Equation 4.6-3 to calculate Abed:

                   Abed = Qe,a/Ue

                   Calculate Qe/a using Equation 4.6-4:

                   Qe,a = Qe KTe + 460)/530]

                   Qe a = _ acfm

                   Assume Ue = 100 ft/sec

                   Abed = - ft2

                b.  Use Equation 4.6-5 to calculate Dbed:

                   Dbed = 1.13 (Abed)a5
                c. Use Equation 4.6-6 to calculate volume of carbon per bed:

                  Vcarbon = (Creq/N)/Pbed

                  Assume pbed = 30 Ib/ft3

                  Vcarbon ~ - ft
                                                            743

-------
                d. Use Equation 4.6-7 to calculate Zbed:

                   ^•bed ~ vcarbon/Abed
                Note: If Qe is greater than about 20,000 scfm, three or more carbon beds
                      may need to be used.

4.6.4.3 Steam Required for Regeneration
                a.  Use Equation 4.6-8 to calculate steam requirements:

                   Qs - [St X Creq/(ereg - 6dry-cool)]/60

                   Assume 0dry-cooi = °-25 nrs-

                   Qs =	Ib/min

                b.  Use Figure 4-14:

                   Qs =	Ib/min

                   Calculate Qs/Abed:

                   Qs/Abed =	Ib steam/min-ft2

                   If Qs/Abed is greater than 4 Ib steam/min-ft2, fluidization of the carbon
                   bed may occur.

4.6.4.4 Condenser
                a.  Use Equation 4.6-10 to calculate H|oad:

                   Hload = 1.1  X 60 X Qs [X  + Cpw(Tsti - Tgto)]

                   Obtain X  and Cpw from  reference 19 based on the values assumed
                   forTstiandTsto.

                   Hload =	Btu/hr

                b.  Use Equation 4.6-9 to calculate Acon:

                   ACon  = H|oad/UATLM

                   Assume U  = 150 Btu/hr-ft2-°F if no other data are available.

                   AT   _ I    ' ' Sti " ' WO'  '  ' StO " ' Wi'   |
                            lnl(Tsti-Two)-

                   144

-------
                  where Twi = 80°FandTwo = 130°F.

                  ATLM = - °F

                  A   -           ft2
                  AACOn — - I I

                c. Use Equations 4.6-11 and -12 to calculate Qw:

                  QCOO!,W ~ H|oacj/[Cpw(Two - Twi)]


                  QCOOI.W =
,,w
                  Qw = 0.002 Qcoo,,

                  Qw = _ gal/min


4.6.4.5 Recovered Product
                a. Use Equation 4.6-13 to calculate Qrec:

                  Qrec = 1.55x10~9xQexHAPexRExMWHAp

                  Q   = _ Ib/hr
                    rec
4.6.5 Evaluation of Permit Application
Compare the results from the calculations and the reported values using Table 4-8.

If the calculated values of Creq, Dbed, Zbed, Qs, Acon, Qw, and Qrec, are different from the
reported values, the differences may be due to the assumptions involved in the calcula-
tions. Discuss the details  of the design and operation of the system with the applicant.

If the calculated values agree with the reported values, then the design and operation of
the proposed carbon adsorber system may be considered appropriate based on the
assumptions made in this handbook.
                                                           745

-------
Table 4-8.  Comparison of Calculated Values and Values Supplied by the Permit
          Applicant for Carbon Adsorption

                                           Calculated    Reported
                                             Value         Value

           Carbon requirement, Creq

           Bed diameter, Dbed

           Bed depth, Zbed

           Steam rate, Qs

           Condenser surface area, Acon

           Cooling water rate, Qw

           Recovered product, Qrec
                  146

-------
                                 Appendix C. 7
                        Calculation Sheet for Absorption

4.7.1 Data Required
       HAP emission stream characteristics:

       1. Maximum flow rate, Qe =	scfm

       2. Temperature, Te =	°F

       3. HAP =	

       4. HAP concentration, HAPe =	ppmv

       5. Pressure, Pe =	mm Hg

       Required removal efficiency, RE =	%

In the case of a permit review, the following data should be supplied by the applicant:

       Absorption system variables at standard conditions (70°F, 1 atm):

         1. Reported  removal efficiency, REreported =	%

         2. Emission  stream flow rate, Qe =	scfm

         3. Temperature of emission stream, Te =	°F

         4. HAP -	

         5. HAP concentration, HAPe =	ppmv

         6. Solvent used =	

         7. Slope of the equilibrium curve, m =	
        8. Solvent flow rate, Lga| =	gal/min

        9. Density of the emission stream, pG =	Ib/ft3

        10. Schmidt No. for the HAP/emission stream and HAP/solvent systems:

           ScG =	
           ScL =	
           (Refer to Appendix B.9, reference 8, or reference 12 for definition and
           calculation of ScG and ScL)
                                                          147

-------
        11.  Properties of the solvent:
           Density, pL =	Ib/ft3
           Viscosity, (XL =	centipoise
        12. Type of packing used =	
        13. Packing constants:
           a = 	b = 	c = 	d  =
           e = 	Y =	s =	g =	
            r =	
        14. Column diameter, DCO|Umn =	ft
        15. Tower height, (packed) HtCO|Umn =	ft
        16. Pressure drop,  APtotai =	in H20
4.7.3 Determination of Absorber System Design and Operating Variables

4.7.3.1 Solvent Flow Rate

               a. Assume a value of 1.6 for AF.
                  Determine "m" from the equilibrium data for the HAP/solvent sys-
                  tem under consideration (see references 18, 22, and 23 for equilibri-
                  um data).
                  m =
                  Use Equation 4.7-3:
                  Qe =	scf m
                  Gmol = 0.155Qe
                  Gm0| =	Ib-moles/hr

                b. Use Equation 4.7-2:

                  l_moi = 1.6m Gmo,

                  Lm0| =	Ib-moles/hr
                  748

-------
                c. Use Equation 4.7-4:

                  Lga, = 0.036 Lmo,

                  Lgai =	gal/min

4.7.3.2 Column Diameter
                a. Use Figure 4-17:

                  Calculate the abscissa (ABS):

                  MWSO|Vent =	Ib/lb-mole

                  L - Lmo, x MWSO|vent

                  L =	Ib/hr

                  MWe =	Ib/lb-mole

                  G  = Gmol x MWe

                  G  =	Ib/hr
                  PG =	Ib/ft3 (refer to Appendix B.9, reference 8, for calcu-
                                        lating this variable)

                  pL =	Ib/ft3 (from reference 18)

                  ABS = (L/G)(PG/PL)°-5

                  ABS =	
                b. From Figure 4-17,  determine the value of the ordinate  (ORD) at
                  flooding conditions.

                  ORD =	
                c. For the type of packing used, determine the packing constants from
                  reference 21:
                  a =
                  e =
                  Determine |iL (from reference 18):

                  M-L =	cp
                                                            149

-------
                d.  Use Equation 4.7-8 to calculate Garea/f:

                   Garea,f = ([ORD pGpL gc]/[(a/e3)W)0-2]}0-5
                   Garea/f =	Ib/sec-ft2

                e.  Assume a value for the fraction of flooding velocity for the proposed
                   design:

                   f =	
                   Use Equation 4.7-9 to calculate Garea:


                        ~ '
                   G    = _ Ib/hr-ft2
                f. Use Equation 4.7-10 to calculate the column cross-sectional area:

                  Acolumn =  G/(3,600 Garea)

                  A  ,                 ft2
                  "column    - ' l

                g.  Use Equation 4.7-11 to calculate the column diameter:

                   '-'column =  ' • '3 \Aco|umn)

                   '-'column ~~ - '*

4.7.3.3 Column Height
       a.  Use Equation 4.7-13 or Figure 4-18 to calculate NOG:

          Using Equation 4.7-13:

          HAPe  = _ ppmv

          HAP0  = HAPe (1 - 0.01 RE)

          HAP0  = _ ppmv

          NOG = ln{(HAPe/HAP0)[1 - (1/AF)] + (1/AF)}/[1 - (1/AF)]

          NOG = -

          Using Figure 4-18:

          HAPP/HAPn = _
              Ql I I/-VI Q


                  150

-------
   At HAPe/HAP0 and 1/AF = 1/1.6 =  0.63, determine NOG:

   NOG =	

b. Use Equations 4.7-14,-15, and -16 to calculate HG, HL, and HOG. Determine the
   packing constants in Equation 4.7-15 using Tables B.9-2 and -3, reference 8,
   or reference 12:

   b =	c -	d  =	

   Y =	s -	

   Determine ScG and ScL using Tables B.9-4 and -5, reference 8, or reference
   12:

   ScG =	

   ScL =	
   *-  ~~ L/ACO|urnn

   L" =	Ib/hr-ft2

   (JLL" =	Ib/hr-ft (from reference 18)

   Calculate HG and HL:

   HG  = [b (3,600 Garea)c/(L")d] (ScG)°-5

   HG  =	ft

   HL = Y(L"/jiL")8 (ScL)0'5

   HL =	ft

   Calculate HOG using AF - 1.6:

   HOG = HG + (1/AF) HL

   HOG =	ft


c.  Use Equation 4.7-12 to calculate HtCO|umn:

   Htcolumn = NOG HQG

   H*column = 	'*
                                                     757

-------
        d.  Use Equation 4.7-18 to calculate Httota|-.

                 = HtCO|umn + 2 + (0.25 DCO|umn)
        e.  Use Equation 4.7-19 to calculate WtCO|umn:

           WtCO|umn = (48 DCO|Umn x Httota|) + 39(DCO|umn)

           WtCO|umn - - Ib

        f. Use Equation 4.7-20 to calculate Vpacking:

          "packing  ~ 0./ob(IJco|urnn)  X Mtco|umn

          \/       —           ft3
           •packing    - ' l

4.7.3.4 Pressure Drop Through the Column
        a.  Use Equation 4.7-21 to calculate APa:

           Determine the constants using Table B.9-6, reference 8, or reference 12:

           9 = -

           r  — r

           APa =  g x TO'8 [10(rL"/pL)] (3,600 Garea)2/PG

           APa =  _ Ib/ft2-ft

        b.  Use Equation 4.7-22 to calculate APtota|:

           APtotal  = AP X Htco|umn

           APtotal  = - Ib/ft2

           APtotal (1/5.2) =  - in H20

4.7.4  Evaluation of Permit Application
Compare the results from the calculations and the values supplied by the permit appli-
cant using Table 4-9. If the calculated values are different from the reported values, the
differences may be due to the  assumptions involved in the  calculations. Therefore,
discuss the details of the proposed design  with the applicant.

If the calculated values agree with  the reported values, then the design of the proposed
absorber system may be considered appropriate based on the assumptions made in this
handbook.
                   752

-------
Table 4-9.    Comparison of Calculated Values and Values Supplied by the Permit
            Applicant for Absorption

                                        Calculated       Reported
                                          Value          Value


           Solvent flow rate, Lgai             ...            ...

           Column diameter, DCO|Umn          ...            ...

           Column height, Htco|umn

           Total column height, Httota|

           Packing volume, Vpacking

           Pressure drop, APtotai

           Column weight, WtCO|Umn
                                                          753

-------
                                  Appendix C.8
                        Calculation Sheet for Condensation

4.8.1 Data Required
        HAP emission stream characteristics:

        1. Maximum flow rate, Qe =	scfm

        2. Temperature, Te =	°F

        3. HAP =	

        4. HAP concentration, HAPe =	ppmv

        5. Moisture content, Me =	%

        6. Pressure, Pe =	mm Hg

        Required removal efficiency, RE =	%

In the case of a permit review for a condenser, the following data should be supplied by
the applicant:

        Condenser system variables at standard conditions (70°F, 1 atm):

         1.  Reported removal efficiency, REreported =	%

         2.  Emission stream flow rate, Qe =	scfm

         3.  Temperature of emission stream, Te =	°F

         4.  HAP =	

         5.  HAP concentration, HAPe =	ppmv

         6.  Moisture content, Me =	%

         7.  Temperature of condensation, Tcon =	°F

         8.  Coolant used =	
         9. Temperature of inlet coolant, T(
coolj
        10. Coolant flow rate, QCOO|ant =	'D/hr

        11. Refrigeration capacity, Ref =	tons

        12. Condenser surface area, Acon -	ft2
                  754

-------
4.8.2 Pretreatment of the Emission Stream
Check to see if moisture content of the emission stream is high. If it is high, dehumidifica-
tion is necessary. This can be carried out in a heat exchanger prior to the condenser.

4.8.3 Condenser System Design Variables
The key design variable is the condensation temperature. Coolant selection will be based
on this temperature.

In evaluating a permit application, use Table 4-10 to determine if the applicant's values
for Tcon, coolant  type, and Tcoo(ri are appropriate:

                T   =            °F
                1 con   	 '
                Coolant type =

                ' cool,i ~ 	
If they are appropriate, proceed with the calculations. Otherwise, reject the proposed
design. The reviewer may then wish to follow the calculation procedure outlined below.

4.8.4 Determination of Condenser System Design Variables

4.8.4.1 Estimation of Condensation Temperature
                a.  Use Equation 4.8-1 to calculate Ppartiai:

                   Ppartiai = 760{(1- 0.01  RE)/[1 - (RE x 1Q-8 HAPe)]}HAPex 10'6

                   Ppartiai =	mm Hg

                b.  Use Figure 4-20 to determine Tcon:

                  T    =           °F
                   i con  	 i
4.8.4.2 Selection of Coolant
                Use Table 4-10 to specify the coolant (also see references 18 and 27):

                Coolant =	
4.8.4.3 Condenser Heat Load
        1.  a. Use Equation 4.8-2 to calculate HAPe,m:

             HAPe,m = (Qe/387) HAPex 10'6

             HAPe/m =	Ib-moles/min

           b. Use Equation 4.8-3 to calculate HAP0,m:

             HAP0,m = (Qe/387)[1-(HAPex10-6)][Pvapor/(Pe-Pvapor)]
                                                            755

-------
              where Pvapor = Ppartia|

              HAP0 m = _ Ib-moles/min

           c.  Use Equation 4.8-4 to calculate HAPcon:

              HAPcon = HAPe,m - HAP0,m

              HAPcon = _ Ib-moles/min

        2. a.  Calculate heat of vaporization (AH) of the HAP from the slope of the graph
              IWPvapor)] vs [1/(Tcon + 460)] for the Pvapor and Tcon ranges of interest.
              See Appendix B.10, reference 8, for details.

              AH = _ Btu/lb-mole

           b.  Use Equation 4.8-5 to calculate Hcon:

              Hcon  = HAPcon[AH + CPHAP (Te - Tcon)]

              where CpHAP can be obtained from references 18 and 27.

              Hcon  = _ Btu/min

           c.  Use Equation 4.8-6 to calculate Huncon:

              Huncon -  HAP0(m CpHAP(Te - Tcon)
              Huncon = - Btu/min

           d. Use Equation 4.8-7 to calculate Hnoncon:

              Hnoncon =  [(Qe/387)-HAPe,m] Cpair (Te - Tcon)

              where Cpair can be obtained from Table B.4-1, reference 8.

              Hnoncon =  - Btu/min

        3. a. Use Equation 4.8-8 to calculate H|0ad:

              Hioad =  1-1 x 60 (Hcon  + Huncon +  Hnoncon)

              H,oad = - Btu/hr


4.8.4.4 Condenser Size
                Use Equation 4.8-9 to calculate Acon:

                   756

-------
                 ACOn - H|oad/U ATLM

                 where ATLM is calculated as follows:

                 AT~LM  =  t(Te - TCOO|/0) - (Tcon - Tcoo(;j)]/ln[(Te - TCOO|;0)/(Tcon - TCOO| f)]

                 Assume: TcooU = Tcon-15, and Tcoo,,0 - TcooU =  25°F

                 T     _              or
                 1 cool,i  ~ - r

                 T  ,   =            °F
                 1 cool,o   - r

                 ATLM  =  - °F

                 Assume: U  = 20 Btu/hr-ft2 -°F (if no other estimate is available).

                 A   = _ ft2
                 "con    - ' L

4.8.4.5 Coolant Flow Rate
                 Use Equation 4.8-10 to calculate QCOOiant:

                 '-'coolant ~  ''load' l^Pcoolant '' cool, o " 'cool,i'J

The value for CpCOO|ant for different coolants can be obtained from references 18 or 27. If
water is used as the coolant, Cpwater can be taken as 1  Btu/lb-°F.

                 Cpcoo,ant = - Btu/lb-°F

                 Qcooiant =  - Ib/hr

4.8.4.6 Refrigeration Capacity
                 Use Equation 4.8-11 to calculate Ref:

                 Ref = H,oad/1 2,000

                 Ref = _ tons

4.8.4.7 Recovered Product
                 Use Equation 4.8-12 to calculate Qrec:

                 Q   = 60  x HAP    x MW
4.8.5 Evaluation of Permit Application
Compare the results from the calculations and the values supplied by the permit appli-
cant using Table 4-11. If the calculated values Tcon, coolant type, Qcooiant' Acon, Ref, and
                                                               757

-------
Qrec are different from the reported values for these variables, the differences may be
due to the assumptions involved in the calculations. Discuss the details of the proposed
design with the permit applicant.

If the calculated values agree with the reported values, then the design and operation of
the proposed condenser system may be considered appropriate based on the assump-
tions made in this handbook.
Table 4-11.    Comparison of Calculated Values and Values Supplied by the Permit
             Applicant for Condensation

                                         Calculated       Reported
                                           Value          Value

          Condensation temperature, Tcon      ...            ...

          Coolant type                       ...            ...

          Coolant flow rate, QCOOiant

          Condenser surface area, Acon        ...            ...

          Refrigeration capacity, Ref

          Recovered product, Qrec
                  158

-------
                                  Appendix C.9
                        Calculation Sheet for Fabric Filters

4.9.1  Data Required
        HAP emission stream characteristics:

        1. Flow rate, Qea =	acfm

        2. Moisture content, Me =	% (vol)

        3. Temperature, Te =	°F

        4. Particle mean diameter, Dp =	|xm

        5. SO3 content  =	ppm (vol)

        6. Particulate content  =	grains/scf

        7. HAP content =	% (mass)

In the case of a permit review, the following data should be supplied by the applicant:

        1. Filter fabric material	

        2. Cleaning method (mechanical shaking, reverse air, pulse-jet)	

        3. Air-to-cloth ratio	ft/min

        4. Baghouse construction configuration  (open  pressure, closed pressure,
           closed suction)	

4.9.2 Pretreatment Considerations
If emission stream temperature is not from 50° to 100°F above the dew point, pretreat-
ment is necessary (see Section 3.3.1 and Appendix  B.1). Pretreatment will  cause two of
the pertinent emission stream characteristics to change; list the new values below.

        1. Maximum flow rate at actual conditions, Qe,a =	acfm

        2. Temperature, Te =	°F

4.9.3 Fabric Filter System Design Variables
        1. Fabric Type(s) (use Table 4-12):

                a. 	

                b. 	

                c. 	
                                                           759

-------
        2.  Cleaning Method(s) (Section 4.8.3.2):

                a.  _

                b.  _

        3.  Air-to-cloth ratio, point or range (Table 4-14) _ ft/min

        4.  Net cloth area, Anc:

                Anc = Qe,a / (A/C ratio)

                where:


                Anc = net cloth area, ft2

                Qe/a = maximum flow rate at actual conditions, acfm

                A/C ratio = air-to-cloth ratio, ft/min

                A  = - ' -
                 DC
                Anc = _ ft2

        5.  Gross cloth area, Atc:

                Ate = Ancx Factor

                where:


                Atc = gross cloth area, ft2

                Factor = value from Table 4-15, dimensionless
                Atc = _ ft2

        6. Baghouse configuration
4.9.4  Evaluation of Permit Application
Using Table 4-16, compare the results from this section and the data supplied by the
permit applicant. As pointed out in the discussion on fabric filter design considerations,
the basic design  parameters are generally selected without the involved, analytical
approach that characterizes many other control systems, such as an absorber system
(Section 4.7). Therefore, in evaluating the reasonableness of any system specifications
                  760

-------
on a permit application, the reviewer's main task will be to examine each parameter in
terms of its compatibility with the gas stream and paniculate conditions and with the
other selected parameters. The following questions should be asked:

        1.  Is the temperature of the emission stream entering the baghouse within 50°
           to 100°F above the stream dew point?

        2.  Is the selected fabric material compatible with the conditions of the emission
           stream; that is, temperature and composition (see Table 4-12)?

        3.  Is the baghouse cleaning  method compatible with the selected fabric materi-
           al and its construction; that  is, material type and woven or felted construc-
           tion (see Section 4.9.3.2 and Table 4-13)?

        4.  Will the selected cleaning mechanism provide the desired control?

        5.  Is the  A/C  ratio appropriate for the  application;  that  is, type  of dust and
           cleaning method used (see Table 4-14)?

        6.  Are the values  provided for  the gas flow rate, A/C ratio, and net cloth area
           consistent? The values can be checked with the following equation:


               A/C ratio = -^La-
                           ^nc

               where:

               A/C ratio = air-to-cloth ratio, ft/min

               Qe,a = emission stream flow rate at actual conditions, acfm

               Anc = net cloth area, ft2

        7.  Is  the baghouse configuration appropriate; that is, is it a negative-pressure
           baghouse?
                                                           767

-------
                                 Appendix C. 10
                  Calculation Sheet for Electrostatic Precipitators
4.10.1  Data Required
        HAP emission stream characteristics:

        1.  Flow rate, Qe,a =	acfm

        2.  Emission stream temperature, Te  =
        3.  Particulate content =	grams/scf

        4.  Moisture content, Me =	% (vol)

        5.  HAP content =	% (mass)

        6.  Drift velocity of particles, Ud =	ft/s

        7.  Collection efficiency, CE =	% mass

In case of a permit review, the following data should be supplied by the applicant. The
design considerations in this section will then be used to check the applicant's design.

        1.  Reported collection efficiency =	%

        2.  Reported drift velocity of particles =	ft/sec

        3.  Reported collection plate area =	ft2

4.10.2 Pretreatment of Emission Stream
If  the emission  stream temperature is not from 50° to  100°F above the  dew  point,
pretreatment is necessary (see Section 3.3.1 and Appendix B.1). Pretreatment will  cause
two of the  pertinent  emission stream characteristics to  change; list the new values
below.

        1.  Maximum flow rate at actual conditions, Qe/a =	acfm

        2.  Temperature, Te  =	°F

4.10.3 ESP Design Variables
Collection plate area is a function of the emission stream  flow rate, drift velocity  of the
particles (Table 4-17), and desired  control efficiency. The  variables are related by the
Deutsch-Anderson equation:


                AP =  60%TJ7 x ln |1 - CE>

                  762

-------
                where:

                Ap = collection plate area, ft2

                Qe>a =  emission stream flow rate at actual conditions as it enters the
                       control device, acfm

                Ud = drift velocity of particles, ft/sec

                CE = required collection efficiency, decimal fraction
                     60 x
                Ap = _ ft2

4,10.4  Evaluation of Permit Application
Using Table 4-18, compare the results from this section and the data supplied by the
permit applicant. In evaluating the reasonableness of ESP design  specifications in a
permit application, the  main task will be to examine each parameter in terms of its
capability with the gas stream conditions.

If the applicant's collection plate area is less than the calculated area, the discrepancy
will most likely be the selected drift velocity. Further discussions with the permit appli-
cant are recommended to evaluate the design assumptions and to reconcile any appar-
ent discrepancies.
                                                            763

-------
                                 Appendix C.11
                     Calculation Sheet for Venturi Scrubbers

4.11.1 Data Required
        HAP emission stream characteristics:

        1. Flow rate Qe/a =	acfm

        2. Temperature, Te =	°F

        3. Moisture content, Me =	%

        4. Required collection efficiency, CE =	%

        5. Particle mean diameter, Dp =	  pm

        6. Particulate content =	grains/scf

        7. HAP content =	% (mass)

In the case of a permit review, the following data should be supplied by the applicant:

        1. Reported pressure drop across venturi  =	in H20

        2. An applicable performance curve for the venturi scrubber

        3. Reported collection efficiency =	%

4.11.2 Pretreatment of Emission Stream
If the emission stream temperature is not from 50° to  100°F above the  dew point,
pretreatment is necessary (see Section 3.3.1 and Appendix B.1). Pretreatment will cause
two of the pertinent emission stream characteristics to  change;  list the new values
below:
        1. Maximum flow rate at actual conditions, Qe a =	acfm

        2. Temperature, Te =	°F

4.11.3  Venturi Scrubber Design Variables

4.11.3.1 Venturi Scrubber Pressure Drop
The pressure drop across the venturi (APV) can be estimated through the use of a venturi
scrubber performance curve  (Figure 4-22) and known values for the required collection
efficiency (CE) and the particle mean diameter (Dp).

               APV -	in H2O

If the estimated APV is greater than 80 in H20, assume that the venturi scrubber cannot
achieve the desired control efficiency.
                  164

-------
4.11.3.2 Materials of Construction
Select the proper material of construction by contacting a vendor, or as a lesser alterna-
tive, by using Table 4-20.

                Material of construction	
4.11.4  Sizing of Venturi Scrubbers
Some  performance curves and cost curves are based on the saturated gas flow rate
(Qe s). If Qe s is needed, it can be calculated as follows:

                Qe,s = Qe,a x (Te,s + 460)/(Te + 460)

                where:

                Qes = saturated emission stream flow rate, acfm

                Te s = temperature of the saturated emission stream, °F

Use Figure 4-22 to determine Tes; the moisture content of the emission stream (Me) must
be in units of Ibs H20/lbs dry air.

               Convert Me (% vol) to units of Ibs H20/lbs dry air, decimal fraction:

                (Me/100) x (18/29)  =	Ib H20/lb dry air

                From Figure 4-22:

                Te,s =	°F
                Qe/s = (	) x (	+ 460) / (	+ 460)

                Qe s =	acfm

4.11.5 Evaluation of Permit Application
Using Table 4-21, compare the results of this section and the data supplied by the permit
applicant. Compare the estimated APv and the reported pressure drop across the ven-
turi, as supplied by the permit applicant.

If the estimated and reported values differ, the differences may be due to the applicant's
use of another performance chart, or a discrepancy between the  required and reported
collection efficiencies. Discuss the details of the design and operation of the system with
the applicant. If there are no differences between the estimated and reported values for
APv, the design and operation of the system can be considered appropriate based on the
assumptions employed in this handbook.
                                                           765

-------
Table 4-21.    Comparison of Calculated Values and Values Supplied by the Permit
             Applicant for Venturi Scrubbers
                                             Calculated     Reported
                                               Value        Value


           Particle Mean Diameter, Dp              ...           ...

           Collection efficiency, CE                 ...           ...

           Pressure drop across venturi, APV        ...           ...
                  166

-------
                Appendix C. 12
Capital and Annualized Cost Calculation Worksheet
                                         167

-------
Table C.12-1.    Preliminary Calculations for Capital Cost Algorithm

(1)  Calculation of Duct Diameter, Dduct (in)
        DH,,,t =12     — x
4     Qe.
                               a
                                    2
        where:
        Qe,a = emission stream flow rate at actual conditions, acfm

              = velocity of gas stream in duct, ft/min
             =  12
        If velocity of gas stream in duct is unknown, use 2,000 ft/min; the equation then
        becomes:

        Dduct =  0.3028 (Qe.a)*

        Dduct =  0.3028 ( _ )* = _ in

(2)  Calculation of Stack Diameter, Dstack (in)

              = 12 (*-x  Qfg'a   V
                        X
        where:
        Qfg,a - actual flue gas flow rate, acfm

        Ustack = velocity of gas in stack, ft/min

                     4
        D- 12
        The gas stream velocity in the stack should be at least 4,000 ft/min. If velocity is
        unknown, use 4,000 ft/min; the equation then becomes:
        Dstack = 0.2141

        Dstack = 0.2141 ( - )* - - in

(3) Calculation of Total System Pressure Drop, APt (in H20)

        APt = APduct + APstack + APdevice#1 + APdevice#2 + APdevice#3

        (Note: See Table 5-7 for AP values.)

        APt  = _ + _  + _ +  _  +

        _ in H20

                  768

-------
Table C.12-2.      Estimate of Capital Costs in Current Dollars
          Cost Elements
                                          Figure or
                                         Table Cost
  Escalation Factor
(Current FE/Base FE)
   see Table 5-2
Current Cost
1. Major Equipment Purchase Cost
    Thermal Incinerator
    Heat Exchanger
    Catalytic Incinerator
      Catalyst0, Vcat =
    Carbon Adsorbed
      Carbond, Creq =
    Absorber6
      Packing6, Vpack =
    Condenser'
      Refrigerant'
    Fabric Filter9
      Bags9, Atc =
    ESPh
    Venturi Scrubber'
        Design Factors'
ratnr3 $
rb $
ratnrc $
ft3*
fird $
Ibx
$
H I aHHfirs6 $
k - ft3 Y
$
$
$
ft2*
$
fir' $
tors'
x ( / )
X ( / )
X ( / )
$/ft3 X ( / )
X ( / )
$/lb x ( / )
x ( / )
x ( / )
$/ft3 x ( / )
x ( / )
x ( / )
x ( / )
$/ft2 X ( / )
X ( / )
X ( / )
X
                                       (Thickness Factor)
                                                                                      =  $.
                                                                                         $-
                                                                                         $-
                                                                 (Composition Factor)
                                                                                         $-
                                                                                         $-
                                                                                         $-
                                                                                         $-
                                                                                         $-
                                                                                         $-
                                                                                         $-
                                                                     SUBTOTAL
2. Auxiliary Equipment Purchase Cost
    Ductwork1                      $.

    Fank
    Motor1

    Stack"1



(Fan
x
(Length)

Current Cost)
x ( /
X ( /
. x 0.15
x ( /
) = $
) = $
= $
) = $
                                                                     SUBTOTAL
3.  Pre-Total Purchase Equipment Cost
    Adjustments"
4.  TOTAL Purchase Equipment Cost
                                          Item 1 Subtotal + Item 2 Subtotal
                                          (Item 3) x -0.091
                                          Item 3 + Adjustments
5.  Instrumentation and Controls
6.  Freight and Taxes
7.  TOTAL Purchased Cost
                                          10% of Item 4
                                          8% of Item 4
                                          Item 4 + Item 5 + Item 6
8.  TOTAL CAPITAL COSTS
                                          F°x (Item?); where F =
                                                                              169

-------
                 Footnotes to Table C. 12-2
a Thermal Incinerator: Figure 5-1, includes fan plus instrumenta-
 tion and control costs for thermal incinerators, in addition to
 the major equipment  purchased  cost. Additional  auxiliary
 equipment (ductwork and stack) purchased costs and costs of
 freight and taxes must be added to obtain the total purchased
 cost.

bHeat Exchangers: If the  HAP control system requires  a heat
 exchanger, obtain the cost  from Figure 5-2, escalate this cost
 using the  appropriate factor, and add to the major equipment
 purchased cost.

0 Catalytic Incinerator: Figure 5-3 provides the cost of a catalytic
 incinerator,  less catalyst costs. The "Table" catalyst cost is
 estimated by multiplying the volume of catalyst required (Vcat)
 by  the catalyst cost factor ($/ft3) found on Table 5-3. Catalyst
 costs, all auxiliary equipment  (ductwork,  fan, and stack) pur-
 chased costs and the cost of instrumentation and controls, and
 freight and taxes must be added to obtain the total purchased
 cost.

dCarbon adsorber: Figure 5-4 (packaged carbon adsorber sys-
 tems) includes the cost of carbon, beds, fan and motor, instru-
 mentation and controls,  and a steam  regenerator. Additional
 auxiliary equipment (ductwork and stack) purchased costs and
 costs of freight  and taxes must be added to obtain the total
 purchased cost. Figure 5-5  (custom carbon adsorber systems)
 includes beds, instrumentation and controls, and a steam re-
 generator, less carbon.  The "Table"  carbon cost for custom
 carbon adsorbers is estimated by multiplying the  weight of
 carbon required (Creq) by the carbon cost factor ($/lb) found on
 Table 5-3. Costs of carbon, all auxiliary equipment (duct, fan,
 stack) purchased costs, and freight and taxes must be added to
 obtain the total purchased cost.

e Absorber: Figure 5-6 does not include the cost of packing,
 platforms, and ladders. The cost of platforms and ladders (Fig-
 ure 5-7) and packing must  be  added to obtain the major pur-
 chased equipment cost. The "Table" packing cost is estimated
 by  multiplying the volume of packing  required  (Vpack)  by  the
 appropriate packing cost factor found on Table 5-4. All auxil-
 iary equipment (ductwork, fan, and stack)  purchased costs,
 and costs of freight and taxes must be added to obtain the total
 purchased cost.

1 Condenser  Systems:  Figure 5-8 yields total capital  costs for
 cold water condenser systems. For systems needing refriger-
 ant, the applicable cost from Figure 5-9 must  be  added to
 obtain the total capital costs. In either case, the escalated cost
 estimate is then placed on Line 8, "TOTAL CAPITAL COSTS."

 9 Fabric Filter Systems: Figure 5-10 gives the cost of a negative
  pressure, insulated baghouse. The curve does not include  bag
  costs. The "Table" bag  cost is estimated by multiplying  the
  gross cloth area required  (Atc) by the appropriate bag cost
  factor found on Table 5-5. Bag costs, all  auxiliary equipment
  (duct, fan, and stack) purchased costs, the cost of instrumenta-
 tion  and controls, and freight and taxes must be added to
 obtain the total purchased cost.

h Electrostatic Precipitators: Figure 5-11 provides the cost for an
 insulated ESP. All auxiliary equipment (duct, fan, and stack)
 purchased costs, the cost of instrumentation and controls, and
 freight and taxes must be added to obtain the total purchased
 cost.

' Venturi  Scrubber: Figure 5-12 includes the cost of instrumen-
 tation and  controls in addition to the major  equipment pur-
 chased  cost. This cost curve is based on a venturi  scrubber
 constructed from 1/8-inch carbon steel. Figure 5-13 is used to
 determine  if 1/8-inch steel is appropriate for a given applica-
 tion  (use the higher curve). If thicker steel is required, Figure
 5-14 yields an  adjustment factor for various steel thicknesses;
 this factor  is used to  escalate the cost obtained from Figure
 5-12.  In addition, if stainless  steel  is required (see Section
 4.11.3.2) multiply the  scrubber cost estimate by 2.3 for 304L
 stainless steel or by 3.2  for 316L stainless steel. Costs of all
 auxiliary equipment (ductwork, fan, and stack) and freight and
 taxes must be  added to obtain the total purchased cost.

1 Ductwork:  Figure  5-15 gives  the cost of straight ductwork
 made of carbon steel for various thicknesses, based on  the
 required duct  diameter. Figure 5-16 gives the cost of straight
 ductwork made of stainless  steel  for  various thicknesses,
 based on the required duct diameter. Preliminary calculations
 (duct diameter, see Table C.12-1) are necessary to  estimate
 ductwork costs.

k Fan: Figure 5-17 gives the cost of a fan based on the gas flow
 rate at actual conditions and the HAP control system pressure
 drop (in inches of H20).The applicable fan class is also based
 on the HAP control system pressure drop. Calculation of the
 total system pressure  drop is presented in Table C.12-1.

' The cost of a motor is estimated as 15% of the fan cost.

"Stack: Figure  5-18  gives the cost of a carbon steel stack at
 various stack  heights and diameters. Figure 5-19 gives  the
 price of a  stainless steel stack at various stack heights and
 diameters. Preliminary calculations (stack diameter, see Table
 C.12-1) are necessary to estimate stack costs. For both figures,
 use the curve that best represents the calculated diameter.

n For thermal incinerators,  carbon adsorbers, and venturi scrub-
 bers, the purchase cost curve includes the cost for instrumen-
 tation and  controls. This  cost (i.e., the "Adjustment") must be
 subtracted out to estimate the total purchased equipment cost.
 This is  done by adding  the Item 1 subtotal and the Item 2
 subtotal and  multiplying the  result  by  -0.091.  This  value is
 added to the preliminary total purchased equipment cost to
 obtain the total purchased equipment cost. For all other major
 equipment, the "Adjustment" equals zero.

0 Obtain factor "F" from "TOTAL" line in Table 5-8.
                           770

-------
Table C.12-3.    Preliminary Calculations for Annualized Cost Algorithm

(1) Calculation of Annual Electricity Requirement, AER (Line 5, Table C.12-6)

   a. Fan Electricity Requirement, PER

     FER = 0.0002 (Qfg/a) x AP x MRS

     where:
     Qfg,a = actual flue gas flow rate, acfm

     AP = total HAP control system pressure drop, in H20 (see Table 5-7)

     MRS = annual operating hours, hr
            (Note: Use 8,600 unless otherwise specified.)

     FER = 0.0002 (	) x	x	=	kWh


   b. Baghouse Electricity Requirement, BER

     (Note: Assume 0.0002 kW are required per ft2 of gross cloth area.)

     BER = 0.0002 (Atc) x MRS

     where:
     Ate = gross cloth area required, ft2

     BER = 0.0002 (	) x	=	kWh


   c. ESP Electricity Requirement, EER

     (Note: Assume 0.0015 kWare required per ft2 of collection area.)

     EER = 0.0015(Ap)xHRS

     where:
     Ap = collection plate area, ft2

     EER = 0.0015 (	) x	=	kWh


   d. Annual Electricity Requirement, AER

     AER = FER + BER + EER

     AER =	+	+	=	kWh
                                                          777

-------
(2) Calculation of Capital Recovery Factor, CRF (Line 18, Table C.12-6)

   CRF = [i(l +  i)n]/[(l + i)n-1]

   where:
   i  = interest rate on borrowed capital, decimal fraction
       (Note: Unless otherwise specified use 10%.)

   n = control device lifetime, years (see Table 5-12)

   CRF = [ _ x(1+ _ )(    )]/[(1+  _ )(    )-1]
(3) Calculation of Annual Operator Labor, OL (Line 9, Table C.1 2-6)

   OL = (MRS) (operator hours per shift) / (operating hours per shift)
         (Note: Obtain operator hr/shift value from Table 5-12)

   OL = ( _ ) x ( _ ) / ( _ ) = _ hr

(4) Calculation of Annual Maintenance Labor, ML (Line 11, Table C.12-6)

   ML = (MRS) (maintenance hours per shift) / (operating hours per shift)
         (Note: Obtain maintenance hr/shift value from Table 5-12)

   ML= ( _ )x( _ )/( _ )  =  hr
                  772

-------
Table C.12-4.    Additional Utility Requirements
(1)  Fuel Requirement for Incinerators (Line 1 or Line 2, Table C.12-6)

    (Note: The design sections for thermal  and catalytic  incinerators are  developed
    under the assumption that natural gas is used as the supplementary fuel. Fuel oil
    could be used, however, the use of natural gas is normal industry practice. If fuel oil
    is used, the equation below can be used by replacing Qf with the fuel oil flow rate in
    units of gallons per minute. The resultant product of the equation [gallons of fuel oil
    required] is then used on Line 2 of Table C.12-6.)

    Fuel Requirement = 60 (Qf) x MRS

    where:
    Qf = supplementary fuel required, scfm

    MRS = annual operating hours, hr
          (Note: Use 8,600 hours unless otherwise specified.)

    Fuel Requirement = 60 (	) x	=	ft3

(2)  Steam Requirement for Carbon Adsorber (Line 4, Table C.12-6)

    (Note: Assume 4 Ib of steam required for each Ib of recovered product.)

    Steam Requirement = 4 (Qrec) x HRS

    where:
    Qrec = quantity of HAP recovered, Ib/hr

    MRS = annual operating hours, hr
          (Note: Use 8,600 hours unless otherwise specified.)

    Steam Requirement = 4 (	) x	=	Ib

(3)  Cooling Water Requirement for Carbon Adsorber (Line 3, Table C.12-6)

    (Note: Assume 12 gal of cooling water required per 100  Ibs steam.)

    Water Requirement = 0.48 (Qrec) x MRS

    where:
    Qrec = quantity of HAP recovered, Ib/hr

    HRS = annual operating hours, hr
          (Note: Use 8,600 hours unless otherwise specified.)

    Water Requirement = 0.48 (	) x	=	gal
                                                          773

-------
(4) Absorbent Requirement for Absorbers (Line 3 or 6, Table C.12-6)

   (Note: Assume no recycle of absorbing fluid [water or solvent].)

   Absorbent Requirement = 60 (Lga|) x MRS

   where:
   Lgai = absorbing fluid flow rate, gal/min

   MRS = annual operating hours, hr
          (Note:  Use 8,600 hours unless otherwise specified.)

   Absorbent Requirement = 60 (	) x	=	gal

(5) Water Requirement for Venturi Scrubbers (Line 3, Table C.12-6)

   (Note: Assume 0.01 gal of water required per acf of emission stream.)

   Water Requirement =  0.6 (Qe/a) x MRS

   where:
   Qe a = emission stream flow rate into scrubber, acfm

   MRS = annual operating hours, hr
          (Note:  Use 8,600 hours unless otherwise specified.)

   Water Requirement =  0.6 (	) x	=	gal
                  774

-------
Table C.I2-5.    Estimation of Replacement Parts Annualized Costs	

(1) Annualized Catalyst Replacement Costs (Line 7, Table C.12-6)

   Over the lifetime of a  catalytic incinerator, the catalyst is depleted and  must be
   replaced (assume catalyst lifetime is 3 years):

   Annual Catalyst Cost = (Catalyst Current Cost8) / 3

   Annual Catalyst Cost = (	) / 3 = $	
(2) Annualized Carbon Replacement Costs (Line 7, Table C.12-6)

   Over the lifetime of a carbon adsorber, the carbon is depleted and must be replaced
   (assume carbon lifetime is 5 years):

   Annual Carbon Cost =  (Carbon Current Cost8) / 5

   Annual Carbon Cost =  (	) / 5 =  $	

(3) Annualized Refrigerant Replacement Costs

   Refrigerant in a condenser needs to be replaced periodically due to system leaks;
   however, the loss rate is typically very low. Therefore, assume the cost of refrigerant
   replacement is negligible.

(4) Annualized Bag Replacement Costs (Line 7, Table C.12-6)

   Over the lifetime of a fabric filter system the bags become worn and must be
   replaced (assume bag lifetime is 2 years):

   Annual Bag Cost = (Bag Current Cost3) / 2

   Annual Bag Cost = (	) / 2 = $	
3See Table C. 12-2.
                                                          175

-------
Table C.12-6.      Estimate of Annualized Costs in Current Dollars
     Cost Elements
Units Costs/Factor
Annual Expenditure
    Current Dollars
Direct Operating Costs
   1.  Natural Gasa
   2.  Fuel Oil8
   3.  Water3
   4.  Steam3
   5.  Electricity"
   6.  Solvent8
   7.  Replacement Parts
   8.  Replacement Labor
   9.  Operator Labor*3
  10.  Supervision Labor
  11.  Maintenance Laborb
  12.  Maintenance Materials
  13.  SUBTOTAL
Indirect Operating Costs
  14.  Overhead
  15.  Property Tax
  16.  Insurance
  17.  Administration
  18.  Capital Recovery"
  19.  SUBTOTAL
  20.  CREDITS
NET ANNUALIZED COSTS
 $0.00425 per ft3            x
 $1.025 per gal              x
 $0.0003 per gal            x
 $0.00504 per Ib            x
 $0.059 per kWh            x
 $	pergal0  x
 As applicable (see Table C.12-5)
 100% of Line 7
 $11.53 per hr              x
 15% of Line 9
 $11.53 per hr              x
 100% of Line 11
 Add Items 1  through 12
 80% of Sum of Lines 8, 9,10, and 11
 1% of Total Capital Costd
 1% of Total Capita! Costd
 2% of Total Capital Costd
 (CRF) x Total Capital Costd; where CRF =
 Add Items 14 through 18
 As applicable (see Section 5.2.3)
 Item  13 + Item 19- Item 20
                 _ft3
                 -gal
                 -gal
                 Jb
                 _kWh
                 -gal
$-
$-
$-
$-
$-
                 . hr
                  hr
$-
$-
$-
$-
$-

$-
$-
$-
$-
$-
$-
$-
$-
8 See Table C. 12-4.
b See Tabled 2-3.
0 As applicable.
dTotal Capital Cost from Line 8 of Table C. 12-2.
                                          6U.8. GOVERNMENT PRINTING OFFICE: 1986-646-116x1*061+7
                        776

-------