EPA-450/2-76-028
November 1976
(OAQPS NO. 1.2-067)
                  OAQPS GUIDELINES
        CONTROL OF VOLATILE
            ORGANIC EMISSIONS
                 FROM EXISTING
        STATIONARY SOURCES -
VOLUME I:  CONTROL METHODS
        FOR SURFACE-COATING
                    OPERATIONS
  U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Air and Waste Management
    Office of Air Quality Planning and Standard*
   Research Triangle Park, North Carolina 27711

-------
                                EPA-450/2-76-028
                              (OAQPS NO. 1.2-067)
        CONTROL OF VOLATILE
ORGANIC EMISSIONS FROM EXISTING
        STATIONARY SOURCES -
   VOLUME I: CONTROL METHODS
FOR SURFACE-COATING OPERATIONS
            Emission Standards and Engineering Division
               Chemical and Petroleum Branch
          U.S. ENVIRONMENTAL PROTECTION AGENCY
             Office of Air and Waste Management
            Office of Air Quality Planning and Standards
            Research Triangle Park, North Carolina 27711

                  November 1976

-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers.  Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35) , Research Triangle Park, North Carolina
27711.
This document has been reviewed by the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, and approved
for publication.  Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
                      Publication No. EPA-450/2-76-028
                                   11

-------
                               NOTICE
     Volume I is the first of a continuing series of reports designed
to assist the State and Regional Offices develop regulations for
industries that emit volatile organics.   It contains information on
control schemes now being used, estimates of the cost of several systems
and provides guidance for sampling and analyzing organic emissions.
     We expect Volume II to be issued by the year's end.  It will
contain detailed information on alternatives available for reducing
emissions from five coating industries; can, coil, fabric, paper and
automobile.  (Preprints of the sections for the first four are
available now).  Future volumes will provide guidance on the control
of organic emissions from other industries.  These should be especially
helpful as revisions are made to future State Implementation Plans.
     All questions concerning this series should be directed to
John Haines, Assistant to the Director, Emission Standards and
Engineering Division, FTS 629-5271.
                                  iii

-------
                         Table of Contents
VOLUME I, CONTROL METHODS FOR SURFACE COATING OPERATIONS
GLOSSARY	vi
1.0  INTRODUCTION 	   1
2.0  BACKGROUND INFORMATION 	  10
     2.1  Oxidants and Volatile Organics  	  10
     2.2  Stationary Sources of Volatile Organics 	  11
          2.2.1  Nationwide Emission Estimates  	  11
          2.2.2  Major Northeastern Sources . .  . .	11
     2.3  References	16
3.0  CONTROL TECHNOLOGY 	  17
     3.1  Introduction	17
     3.2  Add-On Equipment  	  17
          3.2.1  Carbon Adsorption  	  17
          3.2.2  Incineration	38
          3.2.3  Condensation	57
          3.2.4  Absorption (Scrubbing) 	  61
     3.3  Process and Material Changes  	  63
          3.3.1  Water-borne Coatings 	  63
          3.3.2  High-solids Coatings 	  70
          3.3.3  Powder Coatings	75
          3.3.4  Hot Melt Formulations	87
          3.3.5  Electrostatic Spray Coating  	  89
          3.3.6  Electron Beam Curing   	90
          3.3.7  Ultraviolet Curing 	  92
     3.4  References	95
                                 iv

-------
4.0  COST OF VOLATILE ORGANIC CONTROL	98

     4.1  Summary	98

     4.2  General Discussion of Costs 	   98

          4.2.1  Factors Affecting Investment and
                 Annual  Cost	99

          4.2.2  Organic Vapor Control  Costs  	  100

     4.3  References	136

5.0  APPROACHES TO DETERMINATION OF TOTAL "NONMETHANE"
     HYDROCARBONS   	  137

     5.1  Summary   	•	137

     5.2  Introduction	137

     5.3  Measurement Approaches  	  139

     5.4  Sampling	141

     5.5  Analysis	151

     5.6  Conclusions	160

     5.7  References	163

-------
                                 GLOSSARY

acf?p          actual cubic feet per minute
»;tu           British thermal units
cm3           cubic centimeters
CH/j.           methane
CO;?           carbon dioxide
CO            carbon monoxide
EPA           U.S. Environmental Protection Agency
°F            degrees Farenheit
rID           flame ionization detector
ft3           cubic feet
g             grams
gal           gallon
GC            gas chromotography
H2            hydrogen
hr            hour
in.           inch
kW            kilowatt
1 b            pound
LEL           lower explosive limit
max.          maximum
Tin           minute
mg            mi 11i gram
mmHg          millimeters of mercury
MS            mass spectrometry
NAAQS         national ambient air quality  standards
NDIR          nondispersive  infrared absorber
NEDS          National Emissions Data  System
NOX           nitrogen oxides
Og            oxygen
pom           parts per million

-------
scfm          standard cubic feet per minute
TRC           The Research Corporation of New England
(TM)          registered trademark
wt            weight
yr            year
                                     vii

-------
                         1.0  INTRODUCTION

1.1  Purpose of This Series of Documents
     Photochemical oxidant control strategies currently rely heavily on
the substitution of organic solvents that are considered to be of relatively
low photochemical reactivity, coupled with add-on control and process and
material changes to reduce emissions of the compounds of higher reactivity.
Recent information indicates that solvent substitution is of only marginal
effectiveness in lowering ambient oxidant levels.  This series of documents
provides guidance on ways to reduce volatile organic emissions irrespective
of photochemical reactivity.  Specific source categories and available
control technology for these sources are discussed in Volume II.  Subsequent
volumes will cover additional source categories.

1.2  Volatile Organic Sources
     Volatile organics are emitted from a variety of anthropogenic sources.
Total nationwide emissions for 1975 were estimated by EPA to be about
31 million tons, of which 19  million tons were from stationary sources.
Evaporation of organic solvents contributed about 44 percent of the total
from stationary sources; the remainder is from petroleum refining and
distribution, industrial processes, and combustion of fuels and wastes.
     Major industrial sources of volatile organics discussed in Volume II
include automobile and light-duty truck assembly, can coating, coil
coating, fabric coating, and paper coating.  Other industrial categories
which may be examined in future volumes are wood furniture manufacturing,
degreasing operations, drycleaning, graphic arts production, tire manu-
facturing, petroleum refining, magnetic tape coating, miscellaneous

-------
printing and coating, metallurgical  processing,  production  and  use  of
synthetic organics such as plastics,  rubbers  and resins;  production of
various high-volume organic chemicals,  pharmaceutical  production,
stationary combustion sources, and food processing.

1.3  Control Technology
     Emissions of organic air pollutants can  be reduced by  (1)  add-on
control devices that either destroy or collect the organic  for  reuse
or disposal and (2) process or material changes  that reduce or  eliminate
the use of organics.
     Today, the principal add-on control devices for the control  of
volatile organics are:
     •   Catalytic and noncatalytic (thermal)  incinerators
     •   Activated carbon and other types of adsorbers
        Liquid scrubbers or adsorbers
     •   Condensers that use refrigeration or  compression
Incineration is the technique most universally applied by industry,
but it usually requires measurable supplemental  fuel.   Incineration,
therefore, is most acceptable where the developed heat can  offset other
fuel or energy needs.  Adsorption, absorption, and condensation techniques
although effective - are limited to exhaust streams with a much narrower
range of process characteristics than is incineration.

-------
     Process and material  changes are the most diverse options  available
to surface coating industries.   Among the available process and material
changes are:
     •  New coating technologies--e.g.  water-borne, high-solids,
        and powder coatings.
     •  Reduction of air ingestion into the gas stream requiring
        treatment
        Curing coatings in an inert gas
        More efficient coating application methods
Although these changes offer great promise, almost each one is  unique.
Consequently the number necessary to meet all product and process
requirements is large, and conversion costs are frequently very high.
Process and material changes, therefore, can often be implemented only
over much longer time periods than those required for installing add-on
devi ces.
     Several factors influence the effectiveness, cost and applicability
of availab-le control devices or techniques to a given source category.
Quite often the characteristics of a particular process or exhaust gas
stream dictate the use of certain control techniques.  Many control methods
are equivalent in reducing pollution but vary in cost.  In the latter
instances, it is assumed that the operator will select the option that
provides the most reduction for the fewest dollars.
     Other less obvious factors that are unique to the control  of organic
emissions influence the selection of a control option.  For example,
virtually all organics are derived from petroleum, and the increasing
cost of crude oil provides considerable economic incentive to both

-------
reduce solvent consumption and maximize recovery for use.   Other
regulatory requirements also can preclude - or dictate - the use of
certain options.   Insurance and occupational  safety requirements that
specify maximum allowable organic concentrations for fire prevention
and operator safety are examples of such regulatory requirements.
Finally, long-term warranties or customer requirements can limit the
scope of material or process changes.   Section 3 presents a discussion
of the control options.

1.4  Economics
     Economic aspects of control include not only the investment and
annual costs applied to control devices or processes, but also the
ability of the affected industries to absorb these costs.   Although the
affordability question is obviously important, it cannot be addressed
in detail because each firm's financial position is unique.  It must,
however, be addressed by any control official who is considering
imposition of a regulation.
     Section 4 presents a discussion of the important cost factors along
with estimates of the cost to install  and operate incineration and carbon
adsorption systems.  Although generalizations are difficult, the data
indicate that,where feasible, carbon adsorption is the most economical
approach for low concentrations (about 100 ppm) of organics.  For high
concentrations (in the range of 25 percent of the low explosive limit),
carbon adsorption is preferred if the recovered organics can be reused
as solvent; otherwise, incineration with heat recovery is preferred for
high concentrations.

-------
     Process and material changes are not discussed in Section  4.   It
is virtually impossible to generalize because of the almost unlimited
variety of materials and formulations that are available.   For  the same
reason, costs are estimated in this volume for only the major add-on
control options applicable, but not for process  or material changes.
These options will be discussed in future volumes when considered feasible
and reasonable, although in some instances less reasonable examples may
be presented to highlight the wide variation in costs that can  be incurred.
Where possible, the cost effectiveness, that is the cost per unit or
organic emission reduction, is presented for each major control option
to allow comparison between options and industries.
1.5  Test Methodology
     Material balance, the most desirable means of quantifying emissions
of volatile organics, is often not practical because of the complexity
of reactions, combustion processes, or the dispersed nature of some
operations.  Where quantification is necessary, a source test will be
required.  To date, EPA has not adopted a general method for quantification
of "total nonmethane hydrocarbons."  However, methods have been drafted
(and in one case formally proposed) for measurement of gasoline vapor
losses from marketing operations.  The latter techniques are not usually
applicable to the surface coating industry, particularly to baking and
curing processes.
     In-stack samplers for total nonmethane hydrocarbons are not presently
available; therefore, extractive sampling is required.  Various techniques

-------
are available - all  of which rely upon absolute cleanliness of equipment



and a minimal elapsed time between sampling and analysis.



     Analysis is normally possible when a single compound  or very limiter



number of similar organics are emitted and the gas stream  properties are



not severe, e.g., gasoline transfer operations.  Unfortunately, these



favorable conditions occur infrequently.   Where known materials are being



emitted, gas chromatography appears to be the most useful  technique.



Where the materials in the gas stream are unknown, techniques which



combust the nonmethane hydrocarbons and quantify the resultant carbon



dioxide appear accurate.   Severe gas conditions (high temperature, for



example) greatly increase the difficulty of both sampling  and analysis.



     Although no test method is universally applicable, testing is usually



possible using one or more of the techniques described in  Section 5.







1.6  Pollutant Definition and Expression of Emission Limits -- In establishing



standards which apply to sources for which specific compliance test methods



have not been established, consideration should be given to expressing



limitations in absolute terms rather than in terms of a reference method.



In this manner, the intent and effect of the standard is clearly established



without requiring detailed knowledge of the characteristics of a specific



analytical technique as applied to a specific  test stream.



     When a limitation is expressed in absolute terms, compliance determina-



tions may be made with the most practical technique suitable to the



specific case.  For screening purposes, a simple tool such as an explosi-



meter may be satisfactory even though its accuracy may not be more than

-------
a factor of two.  At the other extreme, gas chromatography combined
with mass spectrography could be used where emissions are close to legal
limits.  In most applications, methods employing FID detectors, or
total carbon analysis methods would be used for compliance measurements.
In any case, the absolute accuracy of the test technique as applied to
the specific source would need to be considered in determining compliance.
     As described under "Test Methodology" there are many exhaust streams
where it is impractical to apply methods to quantify mass of volatile organic
compounds.  Therefore, consideration should be given to measuring emissions
and expressing standards in terms of carbon (i.e., "xx pounds per hour
measured and expressed as carbon"), rather than in terms of true mass.
For certain source categories, a more representative molecular weight
other than that of carbon could be assigned.  For example, limitations on
petroleum emissions could be expressed as propane, i.e., "measured as
carbon, expressed as C.H  "  It should be noted that expression of emission
limitations in terms of carbon simply broadens the range of potential
compliance techniques; it does not necessitate the use of total carbon
analysis methods.  In fact, applicability and accuracy of the more commonly
used measurement techniques involving FID or IR detectors are not in any
way diminished by the expression of emission limitations in terms of total
carbon.  For any stream where an FID analyzer is acceptable, results may,
by mathematical manipulation, be "measured and expressed as carbon."
The results so expressed are, however, not necessarily an expression
of the true volatile organic mass such as would be determined by material

-------
balance calculations.   An approach based on the use of flame ionization
detectors is presented in "Regulation 3," as amended October 2,  1974,
Bay Area Air Pollution Control District.  A general total  carbon analysis
method as used by the  Los Angeles County APCD is described in "Total
Combustion Analysis:   A Test Method for Measuring Organic  Carbon Dioxide
in a Solvent Effluent  Control Program," Albert E. Salo, William L.  Oaks,
and Robert D.  MacPhee.  Air Pollution Control District, County of
Los Angeles August 1974, and "Determination  of Solvent Vapor Concentrations
by Total Combustion Analysis:  A Comparison of Infrared with Flame
Ionization Detectors," Albert E. Salo, Samuel Witz, and Robert D. MacPhee,
presented at 68th Annual APCD Convention, Boston, Mass., Paper 75-33.2.
     In establishing volatile organic regulations, the definitions must
state clearly which compounds, if any, are to be exempt.     For example,
if the purpose is to prevent the emission of oxidant precursors, it would
be necessary to identify and delete from consideration carbon monoxide,
methane and other carbon compounds that do not contribute  to photochemical
smog.
     In addition, consideration has to be given to the sample collection
procedure.  The normal procedure is to withdraw a sample,  filter ar /
residual particulate and cool, dehumidify or otherwise treat the sample
such that it is compatible with the FID, IR, explosimeter, etc.   In the
most uncomplicated situation, the sample might be withdrawn at ambient
temperature and pressure and passed through a filter prior to analysis
in an explosimeter.  Where the gases are hot and contain organic gases
and vapor and some particulate, the sampling procedure is necessarily more

-------
complex.   A suggested approach for such an operator would be to filter
at an elevated temperature, probably 200 to 250°F but not greater than
the source temperature, if this is necessary to protect the analytical
instrument.  If this means cooling the sample, as from an oven operating
at 500°F, some condensation of high boiling organics may occur at
250°F but this cannot be avoided.  Organics which pass through the
filter would then be measured by a suitable analytical technique and
considered a volatile organic for enforcement purposes.  Thus a
regulation should specify whether filterable organic material is to
be removed before analysis of organics and should delineate temperature
limitations applicable to such filters.

-------
                 2.0  BACKGROUND INFORMATION

2.1  Oxidants and Volatile Organics
     Oxidants are seldom emitted directly into the atmosphere but result
primarily from a series of chemical reactions between organic compounds
and nitrogen oxides in the presence of sunlight.   The factors that
determine the concentration of oxidants formed in the atmosphere include
the amounts and kinds of organic compounds initially present and the rate
at which additional organics are emitted.  The very complex chemical
reactions involved have been the subject of continuing scientific
investigation during the past twenty years, including studies in the
atmosphere, laboratory (smog chamber) studies, and computer simulations
of the oxidant forming process.
     It has been shown in smog chambers that when exposed to a given
amount of radiant energy, organic compounds do not form oxidants at the
same rate.  Given long periods of exposure to radiant energy and sufficient
quantities of nitrogen oxides, however, almost all organic compounds will
form oxidants.  Highly reactive compounds can result in high oxidant levels
within a few hours.  Less-reactive compounds require longer periods of
irradiation to form oxidants.  Investigators theorize that these n:,.terials
may be carried great distances by an air mass before reacting.  Thus,
they may increase oxidant concentrations at a later time, far downwind
of the source.
                               10

-------
2.2   Stationary Sources of Volatile Orgam'cs
2.2.1  Nationwide Emission Estimates --No truly comprehensive inventory
of organic emissions is available today, although a number of estimates
have been made.  The most recent estimates made by EPA are given in
Table 2.1.  These data are obtained from the data file maintained by
the National Air Data Branch of the Office of Air Quality Planning and
Standards.
     The methodology employed to obtain these emission estimates involves
the use of air pollutant emission factors applied to nationwide published
data on industrial production, fuel consumption, motor vehicle use, and
other activity level indicators for particular source types.  When
appropriate, estimates of average pollutant control efficiency for
particular processes are included to calculate controlled emissions.
     It is noted from Table 2-1 that the emissions from organic solvent
usage comprise a substantial portion of the national total organic emissions.
     In Table 2.2 some of the major sources of organic emissions in the
United States are listed.
2.2.2  Major Northeastern Sources — One of the first tasks in rhis study
was development of a reasonably comprehensive list of the major stationary
sources in the Northeast where photochemical oxidants are a major concern.
     Useful hydrocarbon emissions data had been developed in Region I
by the GCA Corporation  Technology Division for the Metropolitan Boston
Air Quality Control  Region (AQCR) and the Rhode Island - Southeastern
Massachusetts AQCR.   Using this and files from the States of Massachusetts
and Connecticut, Table 2-3, a comprehensive list of large (over 100 tons
per year)  sources was developed.
                               11

-------
Table 2.1   NATIONWIDE EMISSION ESTIMATES OF VOLATILE ORGANIC
                COMPOUNDS - 1975 (PRELIMINARY)
Source Category                              Emissions,  10  tons/yr
Transporation                                              11.7
  Highway                                      10.0
  Non-Highway                                   1.7

Stationary Fuel Combustion                                  1.4
  Electric Utilities                            0.1
  Other                                         1.3

Industrial Processes
  Chemicals                                     1.6
  Petroleum Refining                            0.9
  Metals                                        0.2
  Other                                         0.8

Miscellaneous                                              13.4
  Organic solvents                              8.3
  Oil and Gas Production and marketing          4.2
  Solid Waste                                   0.9

Open Buring                                                 1.0
  Forest Wildfires                              0.6
  Forest Managed Burning                        0.2
  Agricultural Buring                           0.1
  Coal Refuse Burnino                           0.1

  Total                                                    31.0
                               12

-------
TABLE 2-2.  TYPICAL SOURCES OF VOLATILE ORGAMICS WITHIN INDUSTRIAL CATEGORIES

       Fuel combustion, burning and solid waste incineration

         Boilers (coal, oil and gas)
         Wood combustion
         Stationary engines
         Agricultural burning
         Solid waste incineration
           (municipal, industrial, and domestic)
         Coal refuse
         Orchard heaters


       Petroleum refining, distribution, and marketing

         Miscellaneous point sources - refineries
         Vacuum distillation
         Process gas combustion
         Crude, gasoline, distillate, naphtha, etc.
           Transfer losses
           Working losses
           Breathing losses
           Refueling losses
       Chemical manufacturing

         Ammonia
         Carbon black
         Charcoal
         Paint, varnish, and printing ink
         Pharmaceuticals
         Synthetic resins, fibers and plastics
         Ethyl benzene
         Ethyl oxide
         Acrylonitrile
Formaldehyde
Ethylene dichloride
Phthalic anhydride
Maleic anhydride
Ethylene
Propylene
Butadiene
Ethane, butane, propane
Benzene, toluene, xylene
       Evaporation of organic solvents

         Degreasing
         Drycleaning
         Graphic arts
         Metal coating
           Auto assembly
           Can manufacturing
           Coil coating
           Appliances
           Machinery
           Commercial products
           Furniture
         Pesticide Manufacture and use
         Vegetable oil manufacturing

        Other  industrial sources

          Wood processes
            Kraft pulping
            Plywood
          Metallurgical processes
            Cast iron foundries
            By-product coke
          Beer and whiskey production
Textile coating and finishing
  Dyeing
  Scouring
  Rubberizing
  Carpet manufacturing
Paper and film coating
  Coated papers
  Water proofing
  Pressure sensitive tapes
  Magnetic tape
Wood finishing
  Furniture
  Plywood and panel coating
Tire manufacturing
  Mineral  processes
   Asphalt
   Fiberglas
   Mineral wool
  Food processing
  Direct firing of meats
  Deep fat frying

-------
     Additional information supplied by Regions II,  III  and V was



generally in the form of "source-categories-considered-to-be-significant'



with little specific data on emission rates.



     Table 2-3 summarizes the information obtained from  EPA's Regional



Offices.  Again, although the data are not necessarily comprehensive



nor all inclusive, the recent emission inventories tend  to confirm that



these sources emit large amounts of volatile  organics.  The listing in



Tables 2-3 is not in order of amount of emissions or priority.
                               14

-------
Table 2-3  MAJOR INDUSTRIAL ORGANIC EMISSION SOURCE
      CATEGORIES IN THE NORTHEASTERN U.S.
         Paper coating
         Fabric coating and finishing
         Automotive assembly
         Shipbuilding and repair
         Degreasing
         Drycleaning
         Graphic Arts Application
         Tire manufacturing
         Can manufacturing
         Coil coating
         Petroleum refining
         Magnetic tape coating
         Steel production
         Various synthetic organics production
            (plastics, rubbers, resins, etc.)
         Pharmaceutical production
                      15

-------
2.3 References for Chapter 2

1.  Impact of New Source Performance Standards on 1985 National  Emissions
    from Stationary Sources. Vol.  1, The Research Corporation of New England
    (TRC).  Prepared for U.S.  Environmental  Protection Agency, Research
    Triangle Park, N.C.  under Contract No.  68-02-1382, October 1975.


2.  Source Assessment:   Prioritizaticn of Air Pollution from Industrial
    Surface Coating Operations.  Monsanto Research Corporation, Dayton, Ohio.
    Prepared for U.S. Environmental  Protection Agency, Research  Triangle
    Park, N.C. under Contract No.  68-02-1320 (Task 14). Publication
    No. 650/2-75-109-a.   February  1975.
                                     16

-------
                       3.   CONTROL TECHNOLOGY
3.1  INTRODUCTION
     Volatile organic emissions can be reduced by add-on control  devices
and by process and material changes.   This section reviews the most commonly
used methods; important technical aspects of each method are discussed and
problems and limitations considered.

3.2  ADD-ON EQUIPMENT

3.2.1  Carbon Adsorption
3.2.1.1  Introduction — Carbon adsorption uses a physical phenomenon to
separate organic vapors from a gas stream and to concentrate these vapors to
a more manageable form.  The theory of carbon adsorption is discussed,
variables affecting carbon adsorption explored, and design and operation
discussed along with problem areas.
        Carbon adsorption  is applicable to most organic-emitting industries
studied (with a few solvents excepted) but the costs and difficulties will
vary with the specific industry.   A more complete discussion of applicability
in specific industries is  given in additional volumes.
        The term "sorption" applies to two types of phenomena:   (1) where
vapor molecules are concentrated  by adsorption on the surface, and (2) where
vapors are concentrated by absorption of the vapor molecules  into the mass
of the sorbent.  Adsorption is accomplished using four  different types of
materials:   (1) chemically reactive adsorbents (2) polar  adsorbents  (3)
molecular sieves, and  (4)  nonpolar adsorbents.
                                   17

-------
     When adsorption is accompanied by chemical  reaction, the process is
termed "chemisorption," an exothermic process where molecules can only be
one layer thick.  It has been used for odorous sulfur compounds and some
olefins but has little application at this time for organic solvent control.
     When adsorption is not accompanied by chemical reaction, the process
is termed physical adsorption.  In general, polar adsorbents adsorb polar
molecules (e.g., water) preferentially, while nonpolar adsorbents adsorb
nonpolar molecules (e.g., hydrocarbon) preferentially.  Physical adsorption
is less selective than chemisorption, the process is reversible and vapor
molecules can be adsorbed in more than one layer on the surface.  The force
of adsorption is the Van der Waals force.  Molecules of any solid are
attracted to each other and the surface molecules of the adsorption medium
are subject to unbalanced forces that cause vapor or liquid molecules to
be attracted to the surface.  These forces can be induced in such a way
that more than one layer of molecules can be adsorbed.  It must be noted
that in practice, adsorption takes place through a combination of molecular
attraction (Van der Waals forces) and capillary condensation of the vapors
being adsorbed in the pores provided by the extended surface of the adsorbent.
Although activated carbon does not enter into chemical reaction with the
adsorbed vapors, it does catalyze hydrolysis and degradation reactions o"
certain organic solvents, such as ketones.
     Activated carbon is the only physical adsorbent presently in widespread
use for organic vapor collection.  It is a nonpolar adsorbent although it
has some adsorptivity fur water.
                                  18

-------
3.2.1.2  Theory of Carbon Adsorption
     Activated carbon -- Activated carbon can be produced  from a variety
of carbonaceous materials, its characteristics depend on the raw material
and the activation process.   Carbon is activated by oxidation of portions of
the carbon with steam or chemicals.  The end-product of activation is  a
material with a fine, partially interconnected pore structure that has a
very large surface area.
     Primary variables -- The surface area of the activated carbon is  the
primary variable.  The larger the available area, the larger the adsorption
capacity of the carbon, other things being equal.  A typical activated carbon
may have a surface area of 1100 square meters per gram.
     The capacity of carbon is often represented by "adsorption isotherms"
such as Figure 3-1 showing the effect of increasing molecular weight of
organics on carbon capacity.   The isotherms level out as the micropores are
filled.  For pollution control situations, the range of interest is below
a partial pressure of 10 mm of Hg.  The effect of temperature on adsorption
is shown by Figure 3-2.
     The Pol anyi equation can be used for predicting the effect of inlet pol-
lutant concentration on adsorption capacity in the low concentration range.
It can be expressed as follows:
                                                    v
                      Adsorptive Capacity <*•         m
                          g solvent           T log (C /C.)
                        • —	                   o  i
                          g carbon
     Where V  = liquid molar volume of pollutant at normal  boiling point
       T   » absolute temperature
       CQ  = concentration of saturated vapor
       C.|  = initial pollutant vapor concentration into adsorber
                                     19

-------
   0.50
    0.40 —
 °. 0.30
3

a
ui
oa
cc
o
V)
<  0.20  —
o

<
   0.10
    0.00
                         200
300      400       500


  PRESSURE (p), mm Hg
600
700
     Figure 3-1. Adsorption isotherms of hydrocarbon vapors (amount adsorbed, co,

     at pressure, p, on type Columbia L carbon at IOO°F, liquid volume of o> at

     boiling temperaturej.1
800
                                  20

-------
r»
•3
CO
oc
V)
a

-------
no
(\3
               o

               H

               DC
               O
               CJ
                                                  INLET CONCENTRATION

                                                  OUTLET CONCENTRATION BEFORE BREAKTHROUGH
                                                             LENGTH DOWN ADSORBER
                        Figure 3-3.  Movement of vapor concentration distribution curve in carbon bed with increased adsorption time.
                        For curve 4, Ls is the saturated zone and l_z is the adsorption zone.1

-------
This equation shows that low temperature, high V ,  and high concentrations
                                                m
of the organic being adsorbed all tend to increase  capacity.
     Dynamic adsorption -- For a fixed-bed carbon adsorber, the concentra-
tion profile in the bed changes with time as the capacity of the bed is
approached.  Figure 3-3 shows an example of organic vapor concentration
profiles for five different elasped times after regeneration of the carbon.
Curve number 4 represents the "breakthrough" time,  i.e., the time when
organic vapor concentration at the outlet of the adsorber reaches a defined
level (usually 1 percent of the inlet concentration).  In Figure 3-4, the
exit concentration versus time is shown for a specific case.  The break-
through time is 600 minutes and corresponds to Curve 4 of Figure 3-3.  A
time of 1000 minutes corresponds to Curve 5 of Figure 3-3.  The bed should
be regenerated at 600 minutes although full utilization of the bed is not
realized until 1000 minutes.  The length of the bed where active adsorption
occurs is called the adsorption zone.  Length of this zone is usually about
2 inches.
     For exhaust streams containing multiple solvents, vapors of higher V
will displace vapors of lower V .  If the vapors have very different V ,
adsorption  will be as if each solvent was adsorbed independently in a bed
as shown in Figure 3-5.  If the vapors are close in V  there will be co-
adsorption as shown in Figure 3-6.  In either case, the compound with the
lowest Vm will exit the adsorber bed first.  Bed depth for multiple
solvents can be estimated from summation of the bed depths necessary to
adsorb each of the vapors if those vapors were alone in the gas stream.  All
of the adsorption zones must be considered in the calculation of necessary
bed depth.
                                     23

-------
en
                              i
                              "3
                              i
                                  20.0
                                  15.0
10.0
                                                                 10             15

                                                                     ADSORBER LENGTH, in.
                                                           20
25
30
                                   Figure 3-5. Adsorbed vapor profile in activated carbon bed after steady state is established
                                   but with no coadsorption. The odd numbered zones are saturated with the respective sol
                                   vents, and the even numbered zones are the adsorption zones. 1

-------
ro
                                               COMPOUNDE
                                    2	»|-3 -«	4—
                                                               COMPOUNDF
                                                         10             15             20

                                                               ADSORBED LENGTH, in.
25
30
                            Figure 3-6.  Adsorbed vapor profile in activated carbon bed after steady state is established
                            with coadsorption.1

-------
3.2.1.3  Regeneration -- For concentrations greater than a few parts
per million, carbon must be used many times for economic reasons.
To remove adsorbed vapors and reuse the carbon, regeneration is
necessary.  Regeneration is the removal of adsorbed organics from
the carbon and is accomplished by bringing the bed to near equilibrium
at a higher temperature.  Typical regenerants are steam, hot air,
and hot inert gas.  The hotter the regenerant and the longer the
regeneration, the more adsorbed solvent will be removed (desorbed)
from the carbon bed.  There is an economic optimum where adequate
desorption occurs at reasonable energy cost.  The residual solvent in
the bed after regeneration is called the "heel" and "working capacity"
the difference between full capacity and the heel.  Regeneration is
typically about 50 percent complete for each cycle under proper (or
economic) operation.  To optimize the frequence of regeneration, an
automatic device that signals breakthrough may be useful if the size
of the adsorber warrants.
         Steam regeneration -- Steam is the most widely used regenerant.
The bed is closed off from pollutant flow, and steam is introduced
into the bed.  The steam and the pollutant vapors are routed to a
condenser after which they can usually be separated by gravity or
distillation.  Steam regeneration has the advantage of leaving the bed
wet.  By control of the degree of wetness in the bed, various degrees
of gas cooling can be accomplished.  In a variation of this scheme,
steam and pollutant can be incinerated without condensation.
                              27

-------
     Noncondensible gas regeneration -- As inlet concentration decreases,



the bed capacity is reduced.   In order to achieve adequate working capacity



for low concentrations, the heel must be minimized with consequent increased



steam usage.   Figures 3-7 and 3-8 show the effect of vapor concentration



on the steam requirement for regeneration.  The two compounds (propanone



and 4-methyl-2 pentanone) span the range of V  for which carbon adsorption is



applicable.   For concentrations less than 700 ppm, air or inert gas should



be considered for regeneration, especially if (1) the adsorbed solvent has



no  value, (2) the material has appreciable miscibility with water, or



(3) the solvent does not contain large amounts of halogen-, nitrogen- or



sulfur-containing compounds.



     If a noncondensible gas is used for regeneration, the organics can be



removed by condensation, adsorption, and/or incineration.  Condensation of



virtually all organics in a stream is possible if the stream is cooled to a



low enough temperature.  A more practical approach is to condense a portion



of the vapor and to recycle the remainder back through the operating bed.



A schematic of this system is shown in Figure 3-9.



   .  Secondary adsorption of the vapors in a smaller adsorber offers a possible



method of recoving vapors from a dilute source.  The primary adsorber is



regenerated by heated inert gas, yielding a gas stream in which the vapor



concentration is about 40 times as high as in the original stream.  After



cooling, this stream can then be passed through a secondary adsorber which



is regenerated by steam and the organic material recovered.  Reference 1



gives further details fc.  uiis scheme.
                                   28

-------
o
OB
K
O

-------
    1000
o
CO
OC
o
<
a
<
cc
ca
iu
oc
                            212°F STEAM




                            260°F STEAM
     1.0 —
     0.1 —
    0.01
        0        0.1         0.2         0.3        0.4        0.5


                    AMOUNT ADSORBED (co), Ib SOLVENT/lb CARBON



      Figure 3-8. Amount of regenerating agent required to regenerate Gl

      type carbon equilibrated with 4-methyl-2-pentanone at 10 and 3000

      ppm.1
                                       30

-------
IMPURE
 AIR
   ADSORBERS
    RECYCLE
     LINE
                                                               CONDENSER
                 AIR
REGENERATING
     GAS
  GENERATOR
LIQUID SOLVENT
      OR
  POLLUTANT
   Figure 3-9.  Air pollution control system utilizing carbon-resorb with gas
   generation, condensation, and recycle of uncondensed vapor. 1
                              re-
                                 31

-------
3.2.1.4  Problem Areas with Carbon Adsorption -- Several  problems  are
encountered in systems controlling the bake ovens which follow most
surface coating operations.  Thermal  breakdown of the solvents and/or resins
can generate a range of low V  compounds that cannot be adequately adsorbed;
examples are formaldehyde, methanol,  and acetic acid.  Polymerization reactions
may produce tar-like products that will  condense at the operating  temperatures
of carbon adsorbers and not be desorbed, causing fouling.   These complications,
if present, do not make carbon adsorption impossible for ovens, but they will
necessitate precautions or lower carbon life.
     Compounds such as acetone, methyl ethyl Ketone (MEK),  and phenol,
may cause problems because of high heat of adsorption.  With proper design,
however, problems can be avoided.   The main requirement is  the use of a wet
bed and a controlled relative humidity in the inlet gases  to provide a heat
sink for the adsorbed vapors.  Dimethyl formamide (DMF) and nitropropanes
are a more serious problem; carbon adsorption is probably not applicable
where these solvents are used.
     Reuse of solvent -- Unless a single solvent is used and breakdown is
avoided, reuse of the solvent may not be feasible.  Distillation is possible,
but the complexity and cost are so variable that it is difficult to generalize.
Reuse of mixed recovered solvents is  unlikely if the source is a "toll coater"
or "jobber" where many solvents are run on the same machine.  In general, it
has been assumed in this analysis that if the solvent is recovered, it has
fuel value only.
     Particulates -- Particulate matter, if allowed to enter the carbon bed,
can coat the carbon or plug the voids between carbon particles.  Adsorbtivity
                                   32

-------
is decreased and pressure drop increases.   The net result is  that the
carbon must be replaced or cleaned more often.  Si 1 iconized coatings
cause especially difficult problems.
     A solution to a particulate problem is precleaning of the gas.
Fabric collectors, mist eliminators,  or scrubbers may be used.  The
particulate may be very small  in particle size, viscous or tacky and  hence
difficult to remove.  Mist eliminators may be used if the particulate is
a liquid.  If the condensed vapors harden at the operating temperature
of the particulate collector,  frequent cleaning may be necessary.
     Temperature -- Carbon capacity is greater at lower temperatures.
Usually 100°F is considered the maximum entry temperature.  Cooling  may
be accomplished by direct water sprays or by cooling coils.  If condensible
gases are present, a spray cooler and mist eliminator should be placed ahead
of the adsorber if possible.
     Humidity -- Although carbon preferentially adsorbs organic materials,
water will compete with the organics  for adsorption sites.  To minimize this,
relative humidity must be kept below about 50 percent.  A minimum of 20 to
40 percent relative humidity should be maintained, however, especially if
ketones are to be adsorbed.  If gases are hot and wet, cooling followed by
some reheat may be necessary.   Water formed by fuel combustion must be
considered.
     Concentration -- The range of concentrations for which carbon adsorption
is applicable is limited.  The increased operating cost of low concentrations
has been discussed.  There is also a potential problem with high concentrations,
Adsorption is always an exothermic phenomenon; typically 200 to  300 Btu is
generated per Ib of solvent adsorbed.  If sufficient air is not  present to
                                   33

-------
 carry this off, the bed can overheat.  This can result in poor adsorption
 and, in extreme cases, bed fires.  For concentrations over 25 percent of the
 lower explosive limit (LEL),  heating of the bed must be considered in
 calculations.  The problem can be minimized by leaving the bed wet with
 water.
 3.1.2.5  Equipment Design and Operation — Although there are a great variety
 of possible  schemes for carbon adsorption, most applications are similar in
 design.
     The face  velocity is defined as the flow rate divided by cross sectional
 area of the  bed.  At  high face velocities, the pressure drop increases and
 is the controlling factor in practical cases.  Flow velocities for regenerable
 systems vary from 30  to 110 feet per minute.  The sizes of necessary vessels
 are shown  in Table 3-1 assuming a face velocity of 90 feet per minute.  Note
 that for larger beds, design is for horizontal flow.  In evaluating the
 applicability  of carbon adsorption to a source, floor or roof space must
 be considered.

      Table 3-1.   TYPICAL  BED SIZES FOR CYLINDRICAL  CARBON  ADSORBERS

Flow rate,  cfm                                     Size  of  bed

 1.000                                      4 ft diameter - vertical  vessel
 3,000                                      7 ft diameter - vertical  vessel
10,000                                     12 ft diameter - vertical  vessel
30,000                                     12 ft diameter/28 ft long -
                                             horizontal  vessel
 *The lower explosive limit  of a sustance is  the lowest volume percent concentration
  of the vapor in air which  can be ignited at 70°F and normal  atmospheric pressure.
                                  34

-------
The usual practice is to install  at least two adsorbers and operate so that
one is adsorbing while the other is regenerating.   The largest vessel  that
can be factory assembled handles  about 30,000 cfm.   Thus for larger sources
the designer has to choose between multiple packaged units and field
assembled adsorbers.   The materials of construction depend on the source
to be controlled.  If carbon dioxide is present, carbonic acid may be formed.
If halogenated compounds are formed, halogen acids  may form.  Formaldehydes
can yield formic acid.  Often a stainless or high  nickel steel is required.
     Bed depths vary with the organic vapor type,  with the concentration
of organic vapors and with the desired time between regenerations.  The
lower the V , the lower the capacity of the carbon.  Higher concentrations
increase the capacity of the carbon, but also increase the amount of organic
vapor to be adsorbed per unit volume of gas.  The net effect is that at
higher concentrations, the bed must be deeper for a given vapor, face velocity,
and time between regeneration.  Bed depths typically range from 1-1/2 to
3 feet but can be less at low concentrations.  Cycle times for regenerable
systems usually run about 2 hours.  Higher flow rates cause the adsorption zone
to be longer.  For the bed depth range of interest for regenerable systems,
this length of 2 to 4 inches will not be significant in comparison with total
bed depth.
     For a system in which there are no compounds with V  greater than
190 cm /mol, and no polymer formers, or excessive particulates reaching the
carbon, a carbon life of 5 to 10 years can be expected.
                                   35

-------
3.1.2.6  Control Efficiency -- Where carbon adsorption is applicable,
90 percent removal in the carbon adsorber is commonly attainable.
3J.2.7  Adverse Environmental Effects of Carbon Adsorption -- If the
organic solvents to be recovered are miscible with water, a potential water
pollution problem exists.  Ways to avoid this are to treat the water or to
incinerate the desorbed vapor, together with the steam or air purge.
      If incineration is used with solvents containing halogens, sulfur,
or nitrogen compounds, acid gases, SO  and NO  will result.
                                     X       X
3.1.2.8  Effect of Technical Assumptions on Cost Models -- The major technical
assumption made for carbon adsorbers in the calculations in Section 4.2.3.2
was the assumption that the organic vapor was 50 percent benzene and 50 percent
hexane by volume.  The important parameters for these solvents are given in
Table 3-2.
                 Table 3-2.   PROPERTIES OF HEXANE AND BENZENE
Solvent
Hexane
Benzene
Molar volume
(Vm),cnr/mol
140
95
Lower explosive
limit, ppm
12,000
14,000
These compounds are in the middle of the applicable range for carbon adsorption
and are also among the most widely used organic compounds.  If the solvents
used have a higher V , capacity of the bed is increased, with subsequent
increased difficulty of regeneration.  If solvents used  have a  lower V  , capacity
is less and regeneration easier.  Within the applicable  range of V  for carbon
adsorption, costs would not vary greatly from the assumptions given.
                                   36

-------
     The assumption of a hexane/benzene solvent involves  some secondary
assumptions.  Steam can be used for regeneration,  at a fairly low steam
to solvent ratio.  The particular solvents can be  separated easily and
cheaply from the condensed steam by decantation.   Thus, effluent water
contains no measurable organics.  If the solvent contains water soluble
components additional equipment and operations will  be required, such as:
(1) distillation equipment plus water treatment; or (2) incineration of the
desorbed vapor/steam mixture; or (3) hot air regeneration followed by
incineration.  These options can increase costs markedly.  Because solvent
formulations are extremely varied, there is insufficient information upon
which to base reliable estimates of the increased  costs under these conditions.
     Prior to the adoption of Rule 66 solvents with adsorption properties
simi liar to hexane and benzene were quite common.   However, reformulation
led to the widespread introduction of oxygenated components (alcohols, ketones,
esters, and ethers) into most solvent mixtures.  Most of these compounds have
significant solubility in water.  Therefore, a solvent mixture similiar to
benzene/hexane is probably the exception rather than the rule.
     The cost assumptions do not include any particulate removal equipment.
If particulates are present in amounts sufficient to require removal, the
system cost will be increased significantly.
     The cost of steam for desorption is based on  compounds which desorb
readily.  If compounds which are difficult to desorb are encountered, steam
costs will be increased.
                                  37

-------
3.2.2  Incineration



3.2.2.1  Introduction -- Incineration destroys organic emissions by



oxidizing them to carbon dioxide and water vapor.   Incineration is the



most universally applicable control  method for organics; given the



proper conditions, any organic compound will  oxidize.   Oxidation proceeds



more rapidly at higher temperatures  and higher organic pollutant content.



Incinerators (also called afterburners) have been used for many years on



a variety of sources ranging in size from less than 1000 scfm to greater



than 40,000 scfm.



     Use of Existing Process Heaters for Incineration -- The use of



existing boilers and process heaters for destruction of organic emissions



provides for the possibility of pollution control at small capital cost



and little or no fuel cost.  The option is, however, severely limited in



its application.  Some of the requirements are:



     1.  The heater must be operated whenever the pollution source is



         operated; will be uncontrolled during process heater down time.



     2.  The fuel rate to the burner cannot be allowed to fall below



         that required for effective combustion.  On-off burner controls



         are not acceptable.



     3.  Temperature and residence time in the heater firebox must be



         sufficient.



     4.  For proper control, the volume of polluted exhaust gas must be



         much smaller than the burner air requirement and be  located



         close to the process heater.  For most plants doing  surface
                               38

-------
         coating,  especially if surface  coating  is  their main  business,
         the combustion air requirement  is  smaller  than the  coater-
         related exhaust.   In many diversified plants,  the coating
         operation may be  distant from heaters and  boilers.
     5.   Constituents of the coating-related exhaust must  not  damage
         the internals of  the process heater
Few boilers or heaters meet these conditions.
     Use of add-on incinerators -- In noncatalytic  incinerators  (sometimes
called thermal or direct flame incinerators),  a  portion of the polluted
gas may be passed through  the burner(s)  in  which auxiliary fuel  is  fired.
Gases exiting the burner(s) in excess of 2000°F are blended  with the
bypassed gases and held at temperature until reaction is  complete.   The
equilibrium temperature of mixed gases is critical  to effective combustion
of organic pollutants.  A diagram of a typical arrangement is  shown in
Figure 3-10.
     The coupled effect of temperature and residence time is shown  in
Figure 3-11.  Hydrocarbons will first oxidize to water, carbon monoxide
and possibly carbon and partially oxidized organics.  Complete oxidation
converters CO and residuals to carbon dioxide and water.   Figure 3-12
shows the effect of temperature on organic vapor oxidation and carbon
monoxide oxidation.
     A temperature of 1100 to 1250°F at  a residence time  of  0.3 to 0.5
      2
second  is sufficient to achieve 90 percent oxidation of most organic
vapors, but about 1400 to 1500°F may be  necessary to oxidize methane,
cellosolve, and substituted aromatics such as toluene and xylene.2
     Design -- Incineration fuel requirements are determined by the con-
centration of the pollutants, the waste  stream temperature and oxygen

                               39

-------
                         FUME INLET 	
                         CONNECTION    *
   PATH OF FUME FLOW (FUME ITSELF IS
 USED AS SOURCE OF BURNER COMBUSTION
 OXYGEN, ELIMINATING NEED FOR OUTSIDE
AIR ADMISSION AND INCREASED Btu LOAD.)
45,
o
             GAS
         CONNECTION
            PILOT
          ASSEMBLY
                         INCINERATION
                           CHAMBER
                FUME INLET PLENUM
REFRACTORY-LINED
IGNITION CHAMBER
             Figure 3-10.  Typioal burner and chamber arrangement used in direct-flame incinerator.

-------
o
3
QC
a
t-


-------
ro
                                 HYDROCARBONS
                                     ONLY
                                                                   HYDROCARBON AND CARBON
                                                                   MONOXIDE (PER LOS ANGELES
                                                                     AIR POLLUTION CONTROL
                                                                        DISTRICT RULE 66)
                         1150      1200        1250       1300       1350       1400

                                                             TEMPERATURE, °F
1450
1500
1550
                       Figure 3-12. Typical effect of operating temperature on effectiveness of thermal afterburner
                       for destn ction of hydrocarbons and carbon monoxide. 1

-------
level, and the incineration temperature required.   For most organic
solvents, the heat of combustion is about 0.5 Btu/scf for each percent
of the LEL.   This is enough to raise the waste stream temperature about
27.5°F for each percent of the LEL (at 100 percent combustion).   Thus,
at 25 percent of the LEL, the temperature rise will  be 620°F for
90 percent conversion.
     Fuel ~ Natural gas, LPG and distillate and residual  oil  are used to
fuel incinerators.  The use of natural gas or LPG results in lower
maintenance costs; at present, natural gas also is the least expensive
fuel.  However, the dwindling natural gas supplies make it almost a
necessity to provide newly installed incinerators with oil-burning
capabilities.
     In most cases where natural gas or LPG is not available, incinerators
are fixed with distillate fuel oil; residual oil is seldom employed.
Oil flames are more luminous and longer than gas flames, thus require
longer fireboxes.  Almost all fuel oils, even distillate, contain measurable
sulfur compounds.  Residual oils generally have greater sulfur and
particulate contents and many have appreciable nitrogen fractions.
Sulfur oxides, particulates and NO  in combustion products from fuel
                                  X
oil increase pollution emissions and cause corrosion and soot accumulation
or incinerator work and heat transfer surfaces.
     Heat recovery -- Heat recovery offers a way to reduce the energy
consumption of incinerators.  The simplest method is to use the hot
cleaned gases exiting the incinerator to preheat the cooler incoming
gases.  Design is usually for 35 to 90 percent heat recovery efficiency.
                                43

-------
     The maximum usable efficiency is determined by the concentration  of
the orgam'cs in the gases, the temperature of the inlet gases,  and the
maximum temperature that the incinerator and heat exchangers can withstand.
     In a noncatalytic system with a primary heat exchanger, the preheat
temperature should not exceed 680°F, at 25 percent LEL, in order to limit
incinerator exit temperatures to about 1450°F for the protection of the
heat exchanger.  The auxiliary fuel  would heat the stream about 150°F  and
oxidation of the solvent would heat it about 620°F for an exit  temperature
of 680 + 150 + 620 = 1450°F.  At 12 percent LEL the preheat temperature
should not exceed 930°F.  Most burners have not been designed to tolerate
temperatures above 1100°F.
     There are several types of heat recovery equipment using different
materials at various costs.  The most common is the tube and shell heat
exchanger.  The higher temperature exhaust passes over tubes, which have
lower temperature gas or liquid flowing through the tubes; thus increasing
the temperature of that gas or liquid.  Another method uses a rotating
ceramic or metal wheel whose axis is along the wall between two tunnels.
Hot exhaust flows through one tunnel and heats half of the wheel.  Lower
temperature air flows through the other tunnel and is heated as "-.he wheel
rotates.  Another method uses several chambers containing inert ceramic
materials with high heat retention capability.  The hot gas (e.g. from
the incinerator) passes through these beds and heats the ceramic material.
The air flow is then reversed, and lower temperature gas passes through
the heated beds; thus raising the temperature of that gas to near
incineration temperature.   Further details on various heat  recovery
methods and equipment can be obtained from the vendors of incinerators.
                               44

-------
     The use of incinerator exhaust to preheat incinerator inlet air is
often referred to as "primary" heat recovery as illustrated in Case 2 of
Figure 3-13.  Since some systems have a maximum allowable inlet tempera-
ture for the incinerator, it may not be possible to recover all of the heat
available in the incinerator exhaust.  In such case, the inlet to the
incinerator is controlled to minimize fuel requirements.  Note that a non-
catalytic incinerator always requires some fuel to initiate combustion.
     "Secondary" heat recovery uses incinerator exhaust from the primary heat
recovery stage (or from the incinerator directly if there is no primary heat
recovery) to replace energy usage elsewhere in the plant.  This energy can
be used for process heat requirements or for plant heating.  The amount of
energy that a plant can recover and use depends on the individual circum-
stances at the plant.  Usually recovery efficiency of 70 to 80 percent is
achievable, making the net energy consumption of an incinerator minimal or
even negative if gases are near or above 25 percent of the LEL.  The use of
primary and secondary heat recovery is illustrated in Case 3 of Figure 3-13.
It should be noted that heat recovery reduces operating expenses for fuel at
the expense of increased capital costs.  Primary heat recovery systems are
within the incinerator and require no long ducts.  Secondary heat recovery
may be difficult to install on an existing process because the sites where
recovered energy may be used are often distant from the incinerator.  In
applying calculated values for recovered energy values in Case 3 to real
plants, the cost of using recovered energy must be considered.  If secondary
heat recovery is used, often the plant cannot operate unless the control
system is operating because it supplies heat required by the plant.
                                      45

-------
            CASE 1 - BASIC SYSTEM
                  CATALYST, IF ANY
 SOLVENT-
CONTAINING
  OFF
                                            TO
                                        ATMOSPHERE
         FUEL     INCINERATOR
   PROCESS
                                                           CASE 2 - BASIC SYSTEM WITH GAS PREHEAT
                                                                           TO
                                                                        ATMOSPHERE
                                                                                         CATALYST, IF ANY
                                                                             LJW
                                                                             ft
                                                            PREHEATER f_

                                                                                       INCINERATOR
                                                     PROCESS
CASE 3 - PROCESS HEAT RECOVERY WITH GAS PREHEAT
                                                                   CASE 4 - INERT GAS GENERATOR
                                                     COMBUSTION
                                                        AIR      CATALYST, IF ANY
                                                           FUEL     INCINERATOR
                                       ATMOSPHERE
                                                              PROCESS
                                                                                            VENTED TO
                                                                                           ATMOSPHERE
                                                                                      INERT GAS
                        Figure 3-13.  Configurations for catalytic and noncatalytic incineration.

-------
     If the gases in an oven are inert,  that is,  contain  little  oxygen,
explosions are not possible and high  concentrations  of organic solvent
vapor can be handled safely.  The oven exhaust can be  blended with  air
and burned with minimal auxiliary fuel.   The incinerator  may  be  the
source of inert gas for the oven.  Cooling of the incinerator gas  is
necessary, removing energy that can be used elsewhere.  Case  4 of
Figure 3-13 illustrates this scheme.   A modification of the scheme  shown
is the use of an external  inert gas generator.  This scheme can  have  a
significant energy credit because the otherwise discarded organics  are
converted to useful energy.  Because of the specialized nature of  Case  4,
it may not be applicable to retrofits on existing ovens and costs  for this
case are not included in this study.   Note that in  this case  the incinerator
exhaust is in contact with the product.   This limits the available fuel
for this option to natural gas or propane.  The use  of this option would
probably be impossible if any compounds containing  appreciable  sulfur or
halogens are used.
     To illustrate a specific case, Figure 3-14 outlines  a source
controlled by a noncatalytic incinerator.  The source is assumed to
operate 25 percent of the LEL and the incinerator has primary and
secondary heat recovery.  The primary heat exchanger raises the  temperature
to 700°F, at 35 percent heat recovery efficiency.  The heat of combustion
of the organic vapors provides a 620°F additional temperature rise at
90 percent combustion and the burner must supply only enough heat to
raise the gases 80°F to reach the design combustion  temperature  of 1400°F.
Combustion products pass through the primary heat exchanger --  where
                                47

-------
CO
                                                                                          ,ly ATCOMBUSTION = 620°F
                                                                                 ATCOMBUSTIONFUEL'80 F
                                                 PROCESS HEAT RECOVERY
                                   Figure 3~14.  Example of incinerator on oven with primary and secondary heat recovery.

-------
they are cooled to 1025°F -- and enter a 35 percent efficient secondary
heat exchanger.  In the secondary heat exchanger,  further energy is
recovered for use in other areas.  In this example, makeup air for the
source is heated from ambient temperatures to source entrance temperature'-
(higher than oven exit temperatures).
     The energy implications of this scheme can be seen by comparing the
energy input of this controlled source with an uncontrolled source.   In
an uncontrolled source, fuel would be necessary to raise the temperature
of the makeup air from 70°F to 425°F or 355°F.  For a controlled source,
fuel would only need to raise the temperature 800F.  Thus, the energy
input would be reduced by over 80 percent by use of incineration simply
because the organic vapors contribute heat when they burn.
     In the above analysis, the assumptions made are important.  If the
organic vapors are more dilute, the temperature rise due to combustion
will be less.  Heat recovery can be more efficient than 35 percent, making
up for all or some of this difference.  Finally, the analysis assumes
that the heat recovered in the secondary heat exchanger can be used in the
plant.  The heat can be used to produce steam, heat water, supply process
heat or heat buildings.  Obviously, a case-by-case analysis is necessary
to ascertain how much recovered heat could be used.
     Particulates -- The level of particulate concentration found in
surface coating operations should not pose any problems for noncatalytic
volatile organic combustion.  However, an incinerator designed for
hydrocarbon removal usually will not have sufficient residence time to
                                49

-------
efficiently combust organic particulates.
     Safety of preheat -- (At 25 percent of the LEL),  oxidation  rates
at temperatures below 1100°F are slow.   Complete oxidation  can  take
several seconds.  Because the gases are in the heat exchanger for less
than a second preignition should not be a problem using heat recovery
if temperatures are below 1000°F to 1100°F.
     Some problems have occurred in the past with accumulations  of
condensed materials or particulates igniting in the heat recovery devices.
If this occurs, the accumluations  must be periodically removed from the
heat transfer surfaces.  The user should give careful  consideration for
his particular set of circumstances to potential safety problems.  This
is especially true if gases at a high percent of the LEL are preheated.
     Adverse environmental effects -- Sulfur-containing compounds will
be converted to their oxides; halogen-containing compounds  will  be
converted to acids.  A portion of nitrogen-containing  compounds  will be
converted to NOX and additional  NOX will result from thermal fixation.
If use of these compounds cannot be avoided, the benefit from incineration
should be evaluated against the adverse effects and alternate methods
of control should be thoroughly explored.
     The concentration of oxides of nitrogen (NO ) is  about 18 to 22 ppm for
                                                A
natural gas-fired noncatalytic incinerators and 40 to  50 ppm for oil-fire'
noncatalytic incinerators at a temperature of 1500°F,  assuming no nitrogen
containing compounds are incinerated.
                               50

-------
t ff(-> I  <>t Technical Assumptions on Cost Models -- In the cost estimates


i'.tvtioii 4.2.2.1) for noncatalytic incineration, the organic was assumed


to he 50 molar percent hexane and 50 molar percent benzene.  For


noncatalytic incineration, the two important factors are the heat


available per unit volume at the LEL and the temperature necessary for


combustion.  For most solvents, the heat of combustion at the LEL is

                 2
about 50 Btu/scf.   This will vary about +_ 20 percent for almost the entire


range of solvents used (methanol and ethanol are slightly higher).  Thus,


there is little variation due to the type of solvent.


     The assumed temperature of combustion (1400°F) is sufficient to


obtain 95+ percent removal of the entire range of organics used as solvents.


3.2.2.2  Catalytic Incineration -- A catalyst is a substance that speeds up


the rate of chemical reaction at a given temperature without being perma-


nently altered.  The use of a catalyst in an incinerator reportedly enables


satisfactory oxidation rates at temperatures in the range of 500 to 600°F


inlet and 750 to 1000°F outlet.  If heat recovery is not practiced,


significant energy savings are possible by use of a catalyst.  The fuel


savings become less as primary and secondary heat recovery are added.


Because of lower temperatures, materials of construction savings are


possible for heat recovery and for the incinerator itself.  A schematic


of one possible configuration is shown in Figure 3-15.


     Catalysts are specific in the types of reactions they promote.  There


are, however, oxidation catalysts available that will work on a wide range


of organic solvents.   The effect of temperature on conversion for solvent


hydrocarbons is shown in Figure 3-16.  Common catalysts are platinum or


other metals on alumina pellet support or on a honeycomb support.  All-metal


catalysts can also be used.
                                51

-------
                          CLEAN. HOT
                            GASES
CATALYST
ELEMENTS
                                         OVEN
                                         FUMES
                                    PREHEATER
        Figure 3-15.  Schematic diagram of
        catalytic afterburner using torch-
        type preheat burner with flow of
        preheater waste stream through fan
        to promote mixing.!
                     52

-------
en
Co
        K

        LU
        o
        o
                           200
                                            400
                                                                                            1000
                                     600             800



                               TEMPERATURE, °F




Figure 3-16.  Effect of temperature on conversion for catalytic incineration. 1
                                                                                                           1200

-------
     The initial  cost of the catalyst and its periodic replacement
represents, respectively, increased capital  and operating costs.   The
lifetime of the catalyst depends on the rate of catalyst deactivation.
Catalyst Deactivation -- The effectiveness of a catalyst requires the
accessability of "active sites" to reacting molecules.  Every catalyst
will begin to lose its effectiveness as soon as it is put into service.
Compensation for this must be made by either overdesigning the amount of
catalyst in the original charge or raising the temperature into the
catalyst to maintain the required efficiency.  At some time, however,
activity decays to a point where the catalyst must be cleaned or replaced.
Catalysts can be deactivated by normal aging, by use at excessively high
temperature, by coating with particulates, or by poisoning.  Catalyst life-
time of greater than 1 year is considered acceptable.
     Catalyst material can be lost from the support by erosion, attrition,
or vaporization.  These processes increase with temperature.  For metals on
alumina, if the temperature is less than 1100°F, life will be 3 to 5 years
if no deactivation mechanisms are present.  At 1250 to 1300°F, this drops
to 1 year.  Even short-term exposure to 1400 to 1500°F can result in near
total loss of catalytic activity.
      The limited temperature range allowable for catalysts sets  onstraints
on the system.  As mentioned earlier, at 25 percent of the LEL and
90 percent combustion there will be about a 620°F temperature rise as
a result of organic combustion.  Because an  inlet temperature of 500 to
600°F is necessary to initiate combustion, the catalyst bed exit
temperature will be 1120 to 1220°F at 25 percent of the LEL.  This is
                                54

-------
the upper limit for good catalyst life and thus  concentrations  of
greater than 25 percent of the LEL cannot be incinerated in  a catalytic
incinerator without damage to the catalyst.   Restrictions on heat
recovery options are also mandated.   These will  be discussed later.
Coating with particulates — The buildup of condensed polymerized
material or solid particulate can inhibit contact between the active
sites of the catalyst and the gases to be controlled.  Cleaning is the
usual method for reactivation.  Cleaning methods vary with the catalyst
and instructions are usually given by the manufacturer.
Poisoning -- Certain contaminants will chemically react or alloy with
comnon catalysts and cause deactivation.  A common list includes phosphorus,
bismuth, arsenic, antimony, mercury, lead, zinc, and tin.  The first five
are considered fast acting; the last three are slow acting,  especially
below 1100°F.  Areas of care include avoiding the use of phosphate metal
cleaning compounds and galvanized ductwork.  Sulfur and halogens are also
considered catalyst poisons, but their effect is reversible.
Fuel -- Natural gas is the preferred fuel for catalytic incinerators
because of its cleanliness.  If properly designed and operated, a
catalytic incinerator could possibly use distillate oil.  However, much
of the sulfur in the oil would probably be oxidized to SO., which would
subsequently form sulfuric acid mist.  This would necessitate corrosive
resistant materials and would cause the emission of a very undesirable
pollutant.  Therefore, the use of fuel oil (even low sulfur) in a
catalytic incinerator is not recommended.
                                55

-------
Heat Recovery -- The amount of heat that  can  be  transferred  to  the
cooler gases is limited.   The usual  design  is  to have  the  exit
temperature from the catalyst bed at about  1000°F.   If the gas  is at
15 percent of the LEL,  for example, the temperature  rise across the
bed would be about 375°F, and the gas could only be  preheated to about
625°F.  Secondary heat recovery is limited  by the ability  to use the
recovered energy.  If a gas stream is already at combustion  temperature,
it is not useful to use "primary" heat recovery  but  "secondary" heat
recovery may still be possible.  Note that  for catalytic incineration,
no flame initiation is necessary and thus it is  possible to  have no  fuel
input.
      As in noncatalytic systems, heat recovery equipment may need
periodic cleaning if certain streams are to be processed.   For  a discussion
of the safety of preheat, see Section 3.2.2.2.
Adverse environmental effects of catalytic incineration -- As in non-
catalytic incineration, if sulfur- or nitrogen-containing compounds  are
present, their oxides will be generated.   If halogenated compounds  are
present, their acids will be formed.  If it is impossible to avoid  using
these compounds in quantity, incineration may be unwise.
      The concentration of NOX from catalytic incinerators is low, ahout
                     2
15 parts per million,  assuming no nitrogen compounds are incinerated.
Effect of technical assumptions on cost models -- In the cost estimates
for catalytic incineration, the solvent was assumed to be 50 molar percent
hexane and 50 molar percent benzene.  For catalytic incineration, the two
important factors are the heat available per unit volume at the LEL  and
                                                    *
the temperature necessary for catalytic oxidation.
                                56

-------
     As discussed earlier, there is little variation  in  the available
heat from combustion at the LEL.
     The assumed temperature into the catalytic incinerator is  sufficient
to obtain 95 percent removal of the entire range of organics used in
solvents.
3.2.3  Condensation
     Any component of any vapor mixture can be condensed if brought to
equilibrium at a low enough temperature.   The temperature necessary to
achieve a given solvent vapor concentration is dependent on the vapor
pressures of the compounds.
     When cooling a two-component vapor where one component can be
considered noncondensible, for example, a solvent-air mixture,  condensation
will begin when a temperature is reached such that the vapor pressure
of the volatile component is equal to its partial pressure.  The point
where condensation first occurs is called the dew point.  As the vapor
is cooled further, condensation continues such that the partial pressure
stays equal to the vapor pressure.  The less volatile a compound, that is,
the higher the normal boiling point, the lower will be the amount that
can remain vapor at a given temperature.
     In cases where the solvent vapor concentration is high, for example,
from the desorption cycle of a carbon adsorber, condensation is relatively
easy.  However, for sources where concentrations are typically below
25 percent of the LEL, condensation is very difficult.
                                57

-------
     Nonhalogenated organic vapors at 25 percent of the LEL or less are
already very dilute, that is, on the order of 0.15 to 2 percent by
volume.  Figure 3-17 shows the vapor pressure dependence on temperature
for several compounds.  Table 3-3 shows the temperature necessary to
condense various amounts of compounds spanning the volatility range of
compounds used for solvents.  Note that dodecane is not volatile enough
to be used as a major component  in most solvents.  To cool large
quantities of gas from ambient or oven temperatures to below 0°F would
be economically prohibitive.
     The above calculations are for single condensable compound systems.
The calculation methods for multiple condensable component systems are
complex, particularly if there are significant departures from ideal
behavior of the gases and liquids.  As a simplification, the temperatures
necessary for control by condensation can be roughly approximated by the
weighted average of the temperatures necessary for condensation of a
single condensable component system at concentrations equal to the total
organic concentration.
     Totally chlorinated and fluorinated compounds, for example, carbon
tetrachloride and perchlorethylene, are nonflammable and may be handled
safely in all concentrations for nonoccupied areas.  Condensation  , ,iy
be practical if high concentrations of these solvents are present.  In
fact, condensation is widely used in the drycleaning industry for
perch!oroethylene recovery, because the relatively high cost of
chlorinated solvents makes recovery attractive.  Totally chlorinated
compounds, however, are not used extensively in the surface coating industry.
                                 58

-------
              TABLE 3-3. PHYSICAL CONSTANTS AND CONDENSATION  PROPERTIES OF  SOME ORGANIC SOLVENTS
                                     25% of  LEL
                                    concentration
90% condensation
from 25% of LEL
95% condensation
from 25% of LEL
90% condensation
from 200 ppm
en

Compound
Dodecane
Pinene
C10H16
(Terpentine)
0-xylene
C8H10
Toluene
C7H8
Benzene
Methanol
C0H,0
L 0
Hexane
C6H14
Normal
boiling
point, °F
421
300


280
211

175
147

155

(a)
LEL,
0.6
0.7


1.0
1.4

1.3
6.0

1.2

Partial
pressure,
mm of Hg
1.1
1.3


1.9
2.7

2.5
11.4

2.3

Dew
point, °F
120
53


26
5

-15
2

-39

Partial
pressure,
mm of Hg
0.11
0.13


0.19
0.27

0.25
1.14

0.23

(b)
Temp,°F
61
116


-31
-51

-69
-41

-93

Partial
pressure,
mm of Hg
0.55
0.065


0.095
0.135

0.125
0.57

0.115

(b)
Temp,°F
54.4
-31.4


-36.5
-54.3

-96.4
-68.7

-108

Partial
pressure,
mm of Hg
0.15
0.015


0.015
0.015

0.015
0.015

0.015

(b)
Temp,°F
19
-60


-72
-103

-114
-126

-129

   (a)  From Reference 1

   (b)  From Figure 3-16

-------
  1000.0
   10C.O —
E
E

LU
oc


1
     0.1 —
    0.01
       441
141       55.2      -9       -59


       TEMPERATURE, °F
99
-131.7
  Figure 3-17. Vapor pi essures of organic solvents versus temperature.
                                    60

-------
     In summary,  condensation  is  not applicable  as  an  organic solvent
control method for surface coating  except  in  rare  instances  in which
high concentrations of relatively nonflammable materials  are present.
3.2.4  Absorption (Scrubbing)
     Absorption,  as an air pollution control  process,  involves dissolving
a soluble gas component in a relatively nonvolatile liquid.  The  absorption
step is only the  collection step.  After the  gas is dissolved, it must  be
recovered or reacted to an  innocuous form.
     Common absorbents for organic vapors  are water, nonvolatile  organics,
                      3
and aqueous solutions.   Absorption is increased by lower temperatures,
higher solubility of the gas,  higher concentrations of the gas, higher
liquid to gas ratios,  lower concentrations of gas  in the  liquid,  and  greater
contacting surface.  Absorption has been widely  used as a product recovery
step in the petroleum and petrochemical industry where concentrations
are typically very high.   These products are  generally recovered  by
heating to lower the solubility,  or by distillation.
     If a chemical oxidizer is present in the liquid stream, organics
can be oxidized in the stream.  This technique  has been used to  convert
low concentrations of odorous  compounds to less  odorous forms.   The
expense of the oxidizing chemical, however, prevents its  use where
concentrations greater than a few parts per million are present.
     The absorption-regeneration approach for organic solvents  is severely
limited by the low concentrations and consequent low solubilities of most
organic gases in the absorbent.  Exceptions are alcohols, ketones, amines,
glycols, aldehydes, phenol, and organic acids.   Gases may be regenerated
                                 61

-------
by heating and reclaimed by condensation or destroyed by incineration.
     Direct contact with water may be used as a cooling method for
removal of high boiling compounds to avoid opacity problems in the
exhaust or to preclean the air before a carbon absorber, but in most
cases the materials do not go into solution to any appreciable extent.
If water is used for condensation in this way, water treatment may be
necessary before discharge.
     In summary, except for a few specialized cases, absorption is not
applicable to control of organic solvent emissions from surface coating
except as a preliminary step for parti oil ate and high-boiling compound
removal.
                                62

-------
3.3  PROCESS AND MATERIAL CHANGES
3.3.1  Water-borne Coatings
3.3.1.1  Introduction — There is much confusion over the terminology of
coatings containing water as part of their solvent content.  Water-borne,
water-reducible, water-based, water-thinnable, and latex are all used to
describe these coatings.  Strictly speaking, water-borne is the correct
generic term for coatings containing water.   The base of a coating is the
polymer or resin, but many use the term water-based interchangeably with
water-borne.
     There are three types of water-borne coatings:  water-solutions, water-
emulsions, and water-dispersions.  Water-solution coatings feature very
small particles dissolved in a mixture of water and a coupling solvent.
The water-soluble resins normally contain ionizable amine or carboxylic
acid groups that solubilize the molecules.   These systems are more easily
mixed and applied than other water-borne systems.  However, resin properties
that make the resin soluble can also cause water sensitivity after curing
unless additions are made to eliminate this sensitivity.
     Water-emulsions are high molecular weight particles suspended in water
by some stabilizing, dispersing agent.5  The resins, of which vinyls and
acrylics are the most prominent, have very few functional groups and require
emulsifying agents to maintain their form.   Emulsion coatings generally have
the highest water resistance of the water-borne systems.
     Water dispersion coatings are intermediate 1n particle size, in use of
functional  groups, and in water sensitivity.
                                       63

-------
3.3.>.2  Application Techniques — Water-borne coatings may be applied
using any of the methods used for organic solvent-borne coatings, that is,
knife, blade, roller, dip, flow coat, and spray.  The conductivity of water
also enables use of electrophoresis to deposit a coating on conductive
materials.  Conversely, the conductivity makes electrostatic spray more
difficult, although still feasible.  Also, a new dip process 1s available
in which the driving forces are chemical rather than electrical in nature.
     Knife and roller coatings -- Application of water-borne coatings with
a knife or roller presents no special problems.  Corrosion-resistant vessels
and delivery lines must be used as for all water-borne paint facilities.
Small amounts of grease or soot that would be no problem with organic solvent-
borne coatings may produce unacceptable coverage with water-borne coatings.
Therefore, for knife and roller coating and for all other water-borne appli-
caitons, surface cleanliness is vital.
     Dipping — Dipping of materials in water-borne paints has many forms.
A dip Into a bath of water-borne coating may be applicable for a material
that does not need a smooth surface.  This method has been used for parts
of automobiles that are not usually visible after assembly.
     Electrophoretic coating — By using a direct current potential in a
bath and grounding the item to be coated, the item can act as an anode or
cathode and be coated.  This method was first used by Ford Motor Company in
the early sixties.   It is now a proven method for primer application in
                       o
the automotive industry  and is finding use in single coat metal finishing
applications.
                                       64

-------
     This method is applicable for conductive surfaces  and non-conductive
surfaces that can be rendered conductive.   A very even  coat is  produced,
and coverage of edges and hidden parts is  excellent.   The system has
been applied to prime trucks, automobiles, appliances,  and other metal
objects.
     Because the item is immersed, care must be taken to avoid  contamination
of the bath since the bath is very susceptible to impurities.   Washing
with deionized water is necessary before immersion in a dip tank for 1
to 3 minutes.  The thickness of the coating may be adjusted by  the voltage
and to a lesser extent by the immersion time.  The composition  of the
bath fluid is usually 10 to 15 percent solids, 80 to 85 percent water,
and 5 percent organic solvent.  The composition of the coating  as it
emerges from the system is usually above 75 percent solids, and less than
23.5 percent water and 1.5 percent organic solvent.
     The tank must be continuously agitated, filtered,  temperature-, pH-,
and bath solids-controlled to give proper coating.  Cooling must be supplied
to overcome the effects of electrical heating.
     Upon exiting the system, the item must be rinsed to remove excess
coating.  Ultrafiltration is used to recover paint solids from the rinse
water.  The solids are returned to the tank and the pure ultrafiltrate
water is used to supplement the deionized water that is normally used
for the final rinse.

-------
     Control of the system is complex but largely automated.   One operator
normally monitors the equipment and performs analyses on the  bath.   The
operator does not enter the bath area, and hence ventilation  of the tank
is minimal.   During an annual tank cleaning, a spare tank is  used to
hold the coating.
     The electrodeposition system is a signficiant user of electrical
energy.  For the General Motors plant at Framingham, Massachusetts, the
device operates at 1000 amps and 400 volts.  Each automobile  is in the
tank for about 3 minutes; thus, electrical energy usage is about 20 kw/hr
per automobile.  In addition, a 150-ton cooling system is required to
maintain the bath temperature.  This overall energy usage must be balanced
against the energy usage for ventilation and parti oil ate collection in a
spray booth.
     A significant drawback for these systems is their high investment
cost.  The systems are limited to one color per bath and this would be
prohibitive for final coat where many colors were needed.  Lack of gloss
is also a drawback for topcoat application.
     Energy to cure water-borne coatings is considerably higher than that
necessary to cure solvent-borne coatings.  This will be discussed later.
Autodeposited coatings -- A truck plant has recently installed an
"autodeposited" priming system for frames.  '   This plant has been reported
to have now completed 11 months of successful production.  Another company
has recently started operation of a autodeposition line for head light
housings.  A third line is scheduled to start operation very soon.
                                 66

-------
The exact details of the process are proprietary, but it is claimed
that is achieves the same result as electrodeposition without the
electrical requirements.  The process is claimed to now be past the
developmental stage.  Current problems include inability to coat items
of more than one metal (such as carbodies) and the nonsandability of
             9
the coatings.
Flow coating -- Flow coating is simply the immersion of an item in a
coarse spray of coating material.  Most of the coating runs off the item
and is recirculated.  The coating is neither even no'r smooth enough for
visible areas unless sanding is performed on edges.
Spray coating -- Water-borne coatings may be applied by air spray,
airless spray, and electrostatic spray.  Observance of safety considerations
with the high voltage make manual electrostatic spray possible.
     The main problems with spraying of water borne coatings arise from
the physical nature of the solvent.  Organic solvents for coatings are
usually a mixture of several components.  Some solvent evaporates while
the spray is in transit from the gun to the object.  Some evaporates very
quickly to'obtain the viscosity necessary to avoid dripping while allowing
leveling.  The remainder has a higher boiling point and comes ort slowly
to allow leveling of the coating and to avoid bubbles from forming due to
too rapid escape of solvent.  Because there is originally no organic
solvent in incoming air, these evaporation properties are not dependent
on the properties of the incoming air (other than temperature).
                                67

-------
     Water-borne spray coating solvent is typically 70 to 80 percent
water and 20 to 30 percent organic solvent.  The organic solvent is a
necessary part of the coating that gives proper leveling and performance
properties.  Unlike organic solvent mixtures, water is only one compound
with one evaporation rate and boiling point.  The heat of vaporization
is much higher than organic solvents and the rate of evaporation from a
coating is very dependent on the relative humidity of the air surrounding
the coating as well as the cosolvents used.  Roller application followed
immediately by curing has little humidity problem.  When spray coating
with water-borne coatings, humidity control is required.  This increases
energy consumption.  This is an especially severe problem when spray
booths are occupied.
3.3.1.3  Performance and Appearance -- Appearance of water-borne enamels
can be as good as organic solvent-borne enamels if proper curing procedures
are used.  "Orange peel," that is, bumpiness of the surface, is greater
for any enamel than for lacquer.  The organic solvent portion of water-borne
coatings minimizes this "orange peel" effect.  Only a limited number of
resins are available that allow the generation of high-gloss water-borne
coatings.  Water-borne coatings for aluminum are farthest advanced with
tin-plate steel second.  Coaters producing a wide variety of products
and coaters who must warrant products for  long periods  of time in  severe
environments  have the same problems with water-borne coatings as they  do
with other process changes.
3.3.1.4  Energy Consumption  -- The energy  required to remove the solvent
is greater for a given amount of water  than  for  the same amount of organic
solvent.  The heat of vaporization of water  is  about  1000 Btu/lb,  about
                                 68

-------
5 times that for most organic solvents.   The  curing  temperatures  and
time for water-borne coatings is  greater than for organic  solvent-borne
coatings.  It should be noted that the energy for heating  the  part  itself
often exceeds the energy to remove the solvent and cure  the  coating,
particularly with large metal parts.   Counterbalancing these higher
energy items is a significant savings in oven air heating  costs.  Air
recirculation is governed by the  necessity to maintain levels  below
25 percent LEL (or somewhat greater with proper safety controls).
Because of the lower solvent content per unit weight of  solids,  the
volume of exhaust air can be safely reduced somewhat.    In some  cases,
however, the coating quality can  be adversely affected by  too  large a
reduction, because of the drying  properties of the coating.  The net
result is that the energy required to cure water-borne coatings  is
approximately equal to that for organic solvent-borne coatings for  some
applications but will be somewhat higher for most applications.
     If humidity control is required, a significant increase ii.
electrical energy will occur, especially if the coating  is applied  in
an occupied area.  This will be discussed for the automotive and light
truck assembly industry in Volume II.
3.3.1.5  Safety — One of the major advantages of water-borne  coatings
is their non-flammability and low toxicity.  Considerable  savings in
insurance costs can be realized in some cases.

-------
3.3.2  High-solids Coatings
3.3.2.1   Introduction -- The basic ingredient in an organic coating is
the binder or resin.   A resin is a film-forming organic  polymer having
glassy,  plastic, or rubbery properties in the dried state.   As  applied
the resins are liquids of controlled viscosity.  On drying  and  curing
(baking) the materials undergo polymerization and cross-linkage to
form a solid film of the desired properties.
     The materials for resins to be used in conventional  solvent-borne
coatings are "cooked" in resin kettles to yield liquids  which have a
high viscosity at ambient temperatures.  To facilitate compounding with
pigments the resins are dissolved in organic solvents which reduce the
viscosity.  To facilitate application more solvent may be added.  After
application, the solvent evaporates and the resins further polymerize
to yield the solid film.
     The viscosity of the coating as applied can be reduced by using low
molecular weight monomers or "prepolymers," which are applied and then
polymerized (cured) to the high molecular weight solid film.  The amount
of solvent required decreases with decreasing reactant molecular weight.
However, as the molecular weight of the resin formers are reduced, the
difficulty of controlling the polymerization reactions increases.  The
application and curing conditions must be precisely fitted to the reactant
characteristics to yield a film of the desired properties.
     Another method of reducing viscosity of high-solids coatings is
by heating the coating material.  As a rule of thumb an increase in
temperature from 70°F to 125°F is equivalent to a 10 percent solvent
                                 70

-------
reduction.  However, heating can cause loss of solvent crucial  to
the application performance of high-solids coatings.   Heating can also
cause premature gelation of coatings,  particularly on standing.
     The solids content of a coating is expressed as  the volume or
weight of the final cured coating per volume or weight of the coating
as applied.  The term "high solids coatings" is usually reserved for
low solvent coatings which are applied and cured by conventional means.
Low molecular weight materials which are cured by radiation (ultraviolet,
infrared, and electron beam) are classified separately.  Radiation-cured
coatings are discussed in Sections 3.3.3 and 3.3.8.
     High solids coatings were first defined by the Los Angeles County
Air Pollution Control District in its Rule 66; coatings of 80 percent
or more solids (by volume) were exempt from emission limitations.
     Segments of the coatings industry have asked regulatory agencies to
accept 70 percent solids by volume as the definition of high solids.
Because the viscosity of a coating increases rapidly with solids content,
coating manufacturers reported that serious coating application problems
are encountered at 80 percent solids content.  As a consequence, some
regulations have been adopted that give exemptions to coatings with 70
percent solids by volume.  The Boston Transportation Control Plan
is one such regulation.
     Some coatings industry representatives also report that even coatings
of 70 percent solids by volume are not now technologically feasible,
although they consider 70 percent solids by weight as reasonable.
Seventy percent solids by weight corresponds to 54 to 65 percent solids
                                 71

-------
by volume depending on the coating compostion.   Representatives  from
the state agencies of 19 northeastern states (referred to as  the Moodus
Conference) on September 28, 1976, in New York  City recommended  that
coatings with 70 percent solids by weight and coatings whose  liquid
fraction contains no more than 30 percent volatile organic compounds
be allowed emission limits of 550 Ibs per day and 110 Ibs per hour.
3.3.2.2  Materials and Processes — Most high solids resins fall into
two categories, two component ambient temperature cured and single
component heat converted.  The most important types are as follows:

          Two Component                      Single Component
          Ambient Cure                       Heat Converted
          Urethane                              Epoxy
          Acrylic-Urethane                      Acrylic
          Epoxy/amine                           Polyester
                                                Alkyd
     Many two component systems use a catalyst to increase the curing
reactions.  Although these chemical reactions can take place at room
temperature, many plants use low-temperature ovens to cure two-component
systems  rapidly so that the coated product can be handled sooner.  The
oven temperatures required are much lower than for conventional ovens and
the amount of solvent is lower.  This will result in  large energy  savings.
Most thermosetting high-solids coatings are based on  epoxy or urethane
resins.  The most popular two-component coatings are  based on polyurethane
resins.  Coatings properties compare favorably with those obtained from
conventional based enamels.  Toxicity of the isocyanates used for  urethanes
is a potential problem.
                                  72

-------
     Fast-reacting two-component systems are usually applied with
special spray guns that mix the two components  at the spray nozzle.
This equipment, more complicated than conventional  spray equipment,  is
also more expensive.  Some slower-reacting two-component coatings  can
be applied with conventional spray equipment.
     High-solids coatings can be used in a variety of industrial  coating
processes.  Two-component catalytically cured coatings are presently
being air sprayed to coat small metal products.  It might be possible
to coat larger products such as automobiles with such systems.  The coil
coating industry is currently investigating the possibility of using
high-solids coatings, especially two-component coatings.    The can
industry is testing a roll-coat-applied high solids coating for can
exteriors.  Interiors of cans can possibly be coated with spray-applied
high-solids coatings.  Coatings of high viscosity can be applied with a
knife coater, therefore, the paper and cloth industry may be able to
apply high-solids coatings  using existing knife coating equipment.
3.3.2.3  Advantages of High-solids Coatings -- In addition to reduction
of solvent emissions high-solids coatings have other advantages:
     1.  In most cases conventional application methods can be used.
Therefore, conversion costs are low.
     2.  In many cases, the energy required for curing is less than
either conventional solvent coatings or water-borne coatings.  However,
in some cases higher curing temperatures are required and energy usage
is greater than for conventional coatings.
                                  73

-------
     3.   In some cases thick coatings  can be applied,  that  mask  surface
defects  (if desired),  so that less surface preparation for  a  product
is needed.
3.3.2.4   Disadvantages of High-solids  Coatings -- The  limitations  of high-
solids relate to the properties and availability of these coatings:
     1.   Achieving the desired properties in the finished coating  is
difficult.  In conventional coatings the necessary functional properties
are created by polymer building in the resin kettle.  Solvents are then
added to optimize application and appearance.  Most of the  polymerization
in high-solids coatings occurs after application and controlling the
conditions so as to produce the desired properties is  much  more difficult.
     2.   The availability of high-solids coatings is very limited.  These
coatings are just beginning to be converted from laboratory coatings to
proven industrial finishing systems.  Coating manufacturers report that
efforts to produce coatings of 80 percent solids by volume have been
unsuccessful.  Coatings of 70 percent solids are still in the developmental
stage.  Only coatings in the 50 to 60 percent solids range appear to
offer immediate prospects for expansion to widespread usage.
     3.   Pot-life of two component systems is very short, leading to
application difficulties.
     4.   There is a health hazard associated with the isocyanates used
in some two-component systems (urethanes).
                                  74

-------
3.3.2.5  Organic Solvent Emission Reduction Potential -- In order to
compare emissions for coatings of various formulations a common basis
is necessary, such as a given volume or weight of cured solids.  Table 3-4
makes such a comparison for organic solvent-borne coatings and water-borne
coatings.  The water-borne coating is assumed to have a volatile portion
containing 80 percent water and 20 percent organic solvent.  Such coatings
are exempt from emission limitations by Rule 66 type regulations.
     It can be noted that the emissions from a typical water-borne coating
of 25 percent solids by volume will have about the same emission as an
organic solvent-borne coating of 60 percent solids by volume.  Coatings
manufacturers and users have contended that organic solvent-borne coatings
should have the same emission exemption as water-borne coatings, on the
basis of comparable quantities of solids.  They maintain that such an
allowance would greatly stimulate the further development of high-solids
coatings.
3.3.3  Powder Coatings
3.3.3.1  Introduction -- Powder coating involves the application of finely
divided coating solids to a surface, followed by a melting of the coating
solids into a continuous film.  Very little solvent is used  (less than
one percent), and the process is thus almost pollution free.  Several types
of resins may be applied as a powder, but there are limitations on the
type of objects that can be powder coated.  Advantages of and problems
with powder coating are discussed below.
                               75

-------
                     TABLE 3-4.   COMPARISON OF EMISSIONS FROM
                  ORGANIC SOLVENT-BORNE AND WATER-BORNE  COATINGS
             Basis:  1  gal  of solids weighing 11  Ibs.
Organi
Borne
c Solvent-
Coating
Percent Solids
by Vol
12
20
30
40
50
60
70
80
by Wt.
20.5
29.7
41.7
52.7
62.5
72.0
79.7
86.8
Organi
Emissi
gal
7.3
4.0
2.3
1.5
1.0
0.67
0.43
0.25
c
ons
Ibs
42
26
15
9.9
6.6
4.3
2.8
1.6
Water-borne*
Coatings
Percent Solids
by Vol by Wt.
10 13.3
15 19.6
20 26.2
25 31.4
30 36,9
35 42.7
Organic
Emissions
gal Ibs
1.8 11.9
1.1 7.4
0.8 5.3
0.6 4.0
0.5 3.1
0.37 2.4
*Volatile portion is assumed to be 80 percent water and 20 percent organic solvent.
                                        76

-------
3.3.3.2  Advantages of Powder Coating -- In addition to the almost total
elimination of organic solvent emissions, powder coating has several
advantages over solvent-borne coating:
     1.  Single coat application is possible with the fluidized bed
technique for thickness up to 40 mil with one application versus several
applications necessary for solvent-borne coatings.
     2.  Material utilization can approach 100 percent if the powder
can be collected and reused.  This factor allows powder coating to be
potentially the most economical coating material.  The difficulty with
reuse of powder occurs if multiple colors are used.  This will be discussed
later.
     3.  Safety aspects of powder coatings offer some advantages.  Powders
are low in toxicity and nonflammable in storage; however, virtually any
organic powder suspended in air can be explosive.
     4.  Maintenance is generally less because the powder can be vacuumed
from any unbaked areas.  Likewise the paint from any mistakes can simply
be vacuumed off from unbaked items.
     5.  Exhaust air volume is greatly reduced from that used for solvent-
borne spray because application is generally either automatic or else
done in a much smaller area.  Spray booth air theoretically could be
filtered and returned to the plant interior.  Fan power is reduced as are
space cooling or heating requirements.
     6.  Water pollution problems are absent because dry particulate
collection is possible.
                                 77

-------
     7.   Natural  gas usage can be theoretically decreased because little
dilution air is required in ovens.   However,  higher bake temperatures
are usually required, which may result in increased gas usage.
3.3.3.3  Disadvantages of Powder Coating -- Some of the specific problems
with different methods of application are discussed later.  General
problem areas include the following:
     1.  Color change is a difficult problem for powder.  The automobile
and truck assembly industry has this problem in its extreme.  Here,
color changes can occur as often as once a minute and with as little
as 15 seconds to change colors between vehicles.  Furthermore, more than
a dozen colors are usually applied.  For fluidized bed methods, con-
siderable time would be necessary to switch colors because cleanout
of the equipment would be necessary.  A separate dip for each color
would be necessary if color were changed more than once a day.  For
spray operations, the problem of changing colors can be solved by switching
coating supply lines and purging the small amount of powder in the nozzle.
This can, however, be a difficult mechanical problem.  A remaining
difficulty with color change is the problem of reusing overspray.  If
colors become mixed  in the collection device, reuse of powder is impossible
for any applications that change colors more than about once a day,
unless the number of colors are few and it is feasible to use separate
spray areas for each.  Without the ability to reuse the oversprayed
powder, powder coating loses one of its chief economic advantages—
low materials loss.
                                 78

-------
     2.   Color masking is more difficult for electrostatically applied
powder coatings than for solvent-borne coatings.   Fine detail, such  as
printing, is not possible and even two-tone automobiles present a
problem, albeit a solvable one.
     3.   Powder coating materials are discrete particles each of which
must be the same color.  Thus, there can be no user tinting or blending
and all  colors must be available from the manufacturer.  For a coater
that must match a given color, such as in a trademark, the necessary
color may not be available.  Color matching problems, can occur when  using
recycled powder.
     4.,   The high curing temperature required for powder coatings makes
them applicable only for metals and some plastics.
     5.   A typical particle size for sprayed powder coating materials is
                                      12
generally greater than 15 micrometers.    Because 1 mil is about 25
micrometers, it is obvious that thin, uniform spray coatings are difficult
to achieve at coating thickness of less than 2 to 3 mils.  Fluidized bed
coating materials are usually about 200 micrometers in diameter and  thus
are not applicable for thin coatings.
3.3.3.4  Application Methods -- The three general application methods for
powder coatings are electrostatic spray, conventional fluidized bed, and
                            IP
electrostatic fluidized bed.
                                 79

-------
Electrostatic Spray -- In this method, an electrostatic charge is  used
to attract and hold the particles to the object until  they can be  heated
to form a continuous coating.  Electrostatic powder spray coating  involves
the passing of a powder through a spray gun where it is given an electrical
charge.  The object is electrically grounded and the powder is attracted
to the object.  If the object is above the melting point of the powder,
the powder will melt, losing most of its charge and allowing further
attraction.  If the object is not above the melting point, the powder
will be attracted to the surface and the charge will build up because
unmelted powder is generally a poor electrical conductor.  The powder is
attracted to areas with less coating and thus good uniformity of coating
is obtained.  Powder continues to build up until the charge from deposited
powders is such that no more powder is attracted.  The applied powder will
remain attracted to the object until the powder is melted into a continuous,
smooth coating.  Application can be automatic or by manual spray.   An
example of an electrostatic spray line is shown in Figure 3-18.
     In electrostatic spray application, the object to be coated does not
need to be hot and thus does not need to have a high heat capacity.  There
is, however, the requirement that the surface be conductive so that it
may be grounded.  This means that metals or conductive-primed materials
must be used.  There is still a heat-resistence requirement for the object
to be coated since powders require a temperature over 300°F to cure.
This limits the use of powder coatings on paper, fabric, and many types
of plastics.
                                 80

-------
            CURING OVEN
                POWDER
               RECOVERY
               EQUIPMENT
                             Hi,, (
                                           SPRAY GUN
            PREHEAT, IF ANY
POWDER
                                               HIGH VOLTAGE
                                                 SOURCE, IF
                                                   ANY

                                           — COMPRESSED AIR
                           v
          LOAD
Figure 3-18. Typical powder spray operation.


      (Courtesy Society of Manufacturing Engineers)
                   81

-------
     Coverage of recessed areas is better with electrostatic spray
than with fluidized beds since the powder will seek out areas with less
coating.  Preferential  coating of one side of flat objects  can also be
achieved, but some overlap is difficult to avoid.   Masking  is difficult
as the powder can diffuse easily through small cracks or overlap in the
masking.  Finally, some shapes are difficult to coat because of
"Faraday Cage" effect.   This is a phenomenon that prevents  charged
particles from entering certain recessed areas because of repulsion from
charged particles near the area.
     Electrostatic spray coatings have been used for a wide variety of
products including:  metal chairs  , stadium seats  , air filters  ,
telephones  , lawn sprinklers   , appliances   , magnetic wire insulation   ,
and building panels.  General Motors and Ford have experimented with
electrostatic powder spray for automobile bodies.
Conventional Fluidized Bed — If a fluid (such as air) is passed up through
a bed of granular solids with sufficient velocity, the solids will be lifted
by the fluid.  If the vessel design is such that the solids rise and fall
in a confined area, the bed is said to be fluidized because it has many
of properties of a fluid.  The process of fluidization is shown in
Figure 3-19.  Fluidized bed application of powder involves the immersion
of a heated object in the fluidized bed.  Powder adheres to the surface
and may be reheated to create a smoother coating.  A schematic is shown
in Figure 3-20.  To obtain consistent coatings, it is necessary to keep
the part temperature and the bed properties constant.  This is done by
careful control of curing time  and temperature and by frequent addition
of materials to the fluidized bed.
                                 82

-------
                            LOW AIR FLOW;
                           BED STATIONARY
FLUIDIZATION
   BEGINS
                 GAS
                 GAS =
                                                     INCREASING AIR FLOW
                                         Figure 3-19.  Steps in bed fluidization.
                   HIGH AIR FLOW;
                   BEDFLUIDIZED
00
00
                                        -
                                         COATED
                                         OBJECT
                                                  \
                                      GAS
    T
    OBJECT
BEING COATED
                         I
                     ^.FROM
                       OVEN

                     HOTUNCOATED
                        OBJECT
                                                                             !AS DISTRIBUTER
                                          Figure  3-20. Fluidized bed coating process.

-------
     The Huidized bed process can achieve a good quality coating of
                                                             13
7 to 30 mils thickness with greater consistence than dusting.     The
objects to be coated must be capable of being heated to the necessary
temperature and must have sufficient heat capacity to melt the coating.
This essentially limits the method to metal products.  Fluidized bed
coatings do not depend on gravity alone as does dusting, but there are
still difficulties with coating recessed areas or complex shapes.
Fluidized beds have been successfully used in applications such as
dishwasher racks and wire coating.
Electrostatic Fluidized Bed -- A relatively new electrostatic application
method is electrostatic fluidized bed.  The fluidized bed is a source of
charged powder for application.  Electrodes are placed within the fluidized
bed to give the particles a charge.  The charged particles repel each other
and rise in the bed.  As a grounded object is passed over or through the
                                               12
bed, the charged particles are attracted to it.    An oven is then used
to melt the powder to a continuous coating.  Figure 3-21 shows a typical
line.  Again, parts may be preheated to give thicker coatings.
     Some powders, including polyvinyl chloride and nylons, have electrical
                                                   12
properties that make them unusable in this process.    Some of the materials
that can be used are:  epoxies, cellulose acetate butyrates, polyesters,
polypropylenes, polyethylenes, acrylics, f1uorocarbons, and chlorinated
polyethers.

-------
                                                                                       CHARGED POWDER
                                                                                            CLOUD
    AERATED
  POWDER BATH
.UIDIZING
  AIR
FLUIDIZING
   AIR
                                                                                                        HIGH
                                                                                                      VOLTAGE
            Figure 3-21. Typical electrostatic fluidized bed operation.  (Courtesy Society of Manufacturing
            Engineers).
                                                   85

-------
     Other electrostatic application  methods  that  have  had  limited
application are the electrostatic curtain  and electrostatic disc methods.
The electrostatic curtain method uses a traveling  belt  filter  to hold
the powder before spraying.   A jet of air  propels  the powder through  a
charging area towards the object.  This method is  most applicable  for
large flat objects because it applies an equal quantity of  powder  to
                                 13
all areas of the passing surface.    The electrostatic disc method uses a
rotating nonconductive disc as a powder distributer.  The disc is
coated with a resistive material that is charged and imparts a charge
to the powder.  This method has been used to apply powder to steel door
             13
jam sections.
      Dusting  —  Dusting  or  flocking  as  it is  sometimes called, is the
 earliest and  simplest  powder  application method.   It  involves the appli-
 cation  of powdered material  "onto a  surface  which  is  at  a  temperature
 above the melting point  of  the  powder  so as  to  have  the  powder fuse
                                                              12
 upon  hitting  the surface and  form a  coating  over  the  surface.    When
 powder  is applied to an  object  at 350  to 500°F, some  of  it will adhere
 and  become tacky.   The material  may  be reheated to cause the  powder
 to flow into  a smoother coating.   Application may  be  by  automatic or
 manual  means.   Spray guns are normally used  to  distribute  the coating.

      Dusting has  been  used  to apply  powder to raw  rubber goods, to
prevent  offset  in  the  printing  industry, to  color-code steel billets, and
                                              12
to coat  sheet plastic  in  ^rlginal  processing.    Dusting is limited in
the consistency  of coatings that  can be applied.    Application is solely
by gravity settling, so  coating of complex shapes  is difficult if not
impossible.
                                    86

-------
3.3.4  Hot Melt Formulations -- Hot melt  coatings  are  applied  in  a
molten state.  The molten resin film cools soon after  being applied
to the substrate.  Because there is no solvent to  evaporate, virtually
100 percent of the materials that are deposited remain as a solid part
of the coating.  Hot melt coatings are most often  applied to paper,
paperboard, cloth, and plastic.
     When the hot melt coating has been applied and cooled, the film does
not need further heat curing.  Since the  only h^at required is that to
melt the coating initially and to heat the coating applicator, a considerable
energy savings can result compared to oven curing.  Also, because an
oven is not needed, less floor space is needed for the coating line.
The line can be run faster with hot melts than with organic solvent-
borne coatings.  A chilled roll can be used to speed cooling, if necessary.
                                  87

-------
     Hot melt coatings are applied at a variety of temperatures.   Low
melting point coatings are applied at temperatures as  low as 150  to
210°F.  These are materials such as waxes or paraffin  coatings that are
soft and easy to scratch.  To improve scratch resistance, higher  melting
resins are added.  These are usually synthetic organic compounds.  Hot
melt blends with melting points in the range of 300 to 450°F usually
contain no paraffin, waxes, or other low melting point ingredients but
rather are composed of film forming resins and plasticizers.  The resultant
films from such high melting point formulations show properties that
                                                    21
are comparable to high grade solvent-borne coatings.
     Hot melt coatings must, of course, be applied at temperatures that
are higher than the melting point of the coating.  Because the substrate
may be harmed by high temperatures, hot melt coatings with melting points
above 400°F cannot be used for some applications.  However, some
extrusion coatings are heated to 600°F to achieve proper adhesion between
                          22
the polymer and substrate.
     Hot melts may be applied in a variety of ways.  Usually special
heated coating equipment is required.  Lower melting hot melts may be
applied by heated rotogravure or roll coaters.  Extrusion coaters are
widely used also, especially with higher melting point materials.
     Extrusion coatings are a large subclass of hot melt coatings.   In
this type of coating a screw extruder discharges a molten plastic sheet
onto the substrate.  Food containers such as milk carbons are often
coated with extrusion coatings because the plastic  film provdies a good
moisture barrier.

-------
     Ethylene/vinyl acetate copolymer, low and medium density poly-
ethylenes are the resins most widely used for hot melt coatings.
Polyethylene forms a strong film, mixes well  with other resins and
waxes, has good water resistance, has good flexibility at low temperatures,
and is relatively low in cost.  Other resins  used include vinyls,
cellulose esters, alkyl esters, maleic esters, and polystyrenes.   All  of
these materials must have viscosities suitable for application and they
must be chemically stable for long periods in the molten state.
      Hot melts  are applicable to the  paper and fabric  coating  industry,
 although only  for certain  applications.   Thus, hot melt  coatings  cannot
 be judged to be universally  applicable  in the paper  and  fabric coating
 industry at this time.
 3.3.5  Electrostatic Spray Coating  — Electrostatic  spray  coating utilizes
 the attractive  force between materials  of opposite electrical  charge  as an
 aid in applying a uniform  coating to  various  surfaces.   The  method reduces
 overspray and waste and thereby  increases the coatings application
 efficiency over conventional  spray  coating processes.   In  the  case of
 solvent- and water-borne coatings,  this will  in  effect reduce  the amount
 of coating solids and corresponding solvent  carrier  needed for a specific
 coating job.  Electrostatic  spray coating can be used  to apply solvent-
 borne, water-borne, or powder coatings.   Powder  coatings are discussed
 in Section 3.3.3.
      In typical electrostatic spray coating  processes  where  relatively
 nonconductive  solvent-based  coatings  are used, coating particles are
                                  89

-------
                                              23
charged up to 100,000 volts with an electrode.     The grounded object
then attracts the negatively charged particles,  which are captured to
form a film.   In instances where conductive coatings  are used, water-borne
coatings, for example, it is possible to use reverse  polarity, that is,
                                                                   24
charging the  object to be coated and grounding  the spray equipment.
     Electrostatic spray coating is primarily applicable to metal  surface
coating.  It is of particular value for complex shapes.   Glass, plastics,
paper and fabric have been successfully coated with this technique.
Corners or extreme concave shapes on objects may escape  coating due to
                          24
the "Faraday cage" effect.    This phenomenon results from the repulsive
electrical forces in corners or concave areas.
     Electrostatic spray coating has the potential of reducing organic
emissions since it can improve the efficiency of application of solids over
ordinary spray.  This results in less organic solvent emissions.
3.3.6  Electron Beam Curing
3.3.6.1  Introduction -- The electron beam curing process uses high energy
electrons to promote curing of electron beam-curable coatings.  Electrons
bombard a coating to produce free radicals throughout the coating.  This
initiates a crosslinking reaction that continues until the coating is
                                                                     25
cured.  The entire process takes only a few milliseconds to complete.
Since most free radicals are terminated by oxygen.  An inert atmosphere
is desirable so that the surface of the coating will  not be less highly
cross!inked than the interior.
                                  90

-------
3.3.6.2  Energy Consumption -- The energy source for electron beam curing



is electrical.  The electron beam curing unit contains an electron



accelerator that activates the electrons to the necessary energy state.



The high energy electrons are emitted, curing the coating by a stationary



or moving beam.


     The energy requirements  for electron beam curing are dependent on



the size of the unit and the  coating thickness but are typically lower



than for thermal curing.  There is an additional  energy savings



because of the instant startup and shutdown capability of the



electron beam unit.



3.3.6.3  Safety -- Electron beam curing units must be shielded properly



to avoid radiation exposure.   According to occupational Safety and Health



Administration regulations, exposure should not exceed Smillir'erns  of

                                                                oc

radiation in 1 hour and 100 mlllirems in any 5 consecutive days.



     Some electron beam-curable coatings may contain monomers that



are toxic.  Caution should be taken when using such monomers.



3.3.6.4  Organic Solvent Emissions Reduction Potential — There have



been few, if any, tests performed to quantify organic vapors emitted



during the curing process.  It is generally assumed that some low



molecular weight organic compounds are emitted during curing even



though all the components are reactive.  There also may be some ozone


                                         27
generated from the curing process itself.





3.3.6.5 ^Application to Industries Studied -- The use of electron beam



curing is most effecti .- un flat surfaces where the electron beam strikes



the surface vertically.  If the beam strikes the surface at an angle



closer to the horizontal, the amount of absorbed energy can be too small



and the coating will not cure properly.
                                 91

-------
      Electron  beam curing,  unlike  ultraviolet  light  curing,  can  cure
 thick and pigmented coatings  because  of the  penetrating  power  of the
 electrons.
      Because electron beam  curing  uses  relatively new technology,
the coatings necessary for the electron  beam  curing process  are in
the early stages of research and development.   The use of electron
beam curing is very limited  at the  present time.

3.3.7  Ultraviolet Curing
3.3.7.1  Introduction -- In  ultraviolet  curing, ultraviolet  light reacts
with photosensitizers in the coating to  initiate crosslinking to form
a solid film.  The basic components of an ultraviolet curable coating
are:  an ultraviolet-curable base polymer, diluent monomers,  and ultra-
                                28
violet photochemical initiators.
     The ultraviolet-curable polymers  provide most of the desired
coating properties.  The diluent monomers decrease the viscosity of the
polymers, increase the crosslinking density,  and improve  other features
of the coating such as gloss, hardness,  and curing speed.  The photo-
chemical initiators are unstable chemicals that form  free radicals when
                                                                    29
bombarded by ultraviolet light to initiate the crosslinking process.
     The energy source used for ultraviolet curing is electrically
produced ultraviolet light energy such as from mercury vapor lamps.
The use of ultraviolet light for curing is most effective on flat
surfaces where the light reaches the surface  vertically.   When the
ultraviolet light strikes a surface at an angle closer to the horizontal,
the amount of absorbed light can be too small for effective curing.
Obviously, no curing will occur if an area is shielded from the light.
                                 92

-------
3.3.7.2  Performance and Appearance — The actual  performance  and appearance



of ultraviolet-curable coatings is  not only dependent on  the base polymers,



diluent monomers, and photochemical initiators,  but also  on other agents



such as pigments, fillers,  and mar  resistors added to the coating to



provide the desired properties.



     In certain industries, the use of ultraviolet light  curing has been



successful, although this success has been limited mostly to semi transparent



coatings, such as inks.   Ultraviolet cured polyester based coatings have



made a significant penetration into the forest products industry as filler



coatings for particleboard.  Most uses of ultraviolet coatings, however,



are still in the research and development stage.  Major problems are



curing of thick coatings and coatings with pigmentation.   The  main



difficulty with pigmentation is that the pigment particles absorb or



reflect ultraviolet light,  thus reducing the light energy available to



cure the coatings in the deeper layers of the coating.



3.3.7.3  Energy Consumption -- Because little if any flammable solvent is



emitted, the amount of dilution air flow through ovens can be greatly



reduced.  There is a substantial decrease in energy usage compared with



thermal curing.  An ultraviolet curing unit may use only one-third the


                                  30
energy of a standard thermal oven.



3.3.7.4  Safety -- The ultraviolet curing equipment must be shielded



properly to avoid exposure of the equipment operator.  Exposure  at short


                                                          31
distances can cause severe burns to the skin and the eyes.
                                 93

-------
     Certain ultraviolet coating materials may produce skin and eye
irritation.  Others, such as those containing "the more volatile crylic
                                                        31
monomers, are considered toxic and hazardous chemicals."    The handling
of ultraviolet-curable coatings requires care and caution.
3.3.7.5  Organic Solvent Emissions Reduction Potential -- There have been
few, if any, emission tests performed to determine whether volatile
organics are emitted during ultraviolet curing.   Some low molecular
weight organic compounds are probably emitted during the ultraviolet
curing process even though all the components of the coating are reactive,
                                 94

-------
                          3.4  REFERENCES

1.   Package Sorption Systems  Study,  MSA  Corporation,  Evans City, Pa.,
    Prepared for U.S.  Environmental  Protection Agency,  Research Triangle
    Park, N.C.  under Contract EHSD 71-2.   Publication No. EPA  R2-73-202.
    April 1973.
2.   Rolke, R.W.  et al.   Afterburner Systems  Study,  Shell  Development
    Company, Emeryville, Cal., Prepared  for  U.S.  Environmental Protection
    Agency, Research Triangle Park,  N.C.  under Contract No.  ESHD 71-3.
    Publication No. EPA-R2-72-062.  August 1972.
3.   Control Technology for Hydrocarbons  and  Organic Solvent  Emissions
    from Stationary Sources,  U.S.  Department of  Health, Education  and
    Welfare, National  Air Pollution Control  Administration,  Washington,
    D.C.  Publication No. AP-68, March 1970.
4.   Wildman, G.C., and B.G.  Bufkin.   Waterborne's Position  in  the
    Spectrum of Industrial Coatings:  Comparison to Solvent Types  and
    Availability.  American Paint and Coatings Jounral, p.  18-22 and
    56-59.  July 14, 1975.
5.   Anisfield,  j.  Powder's Competition.   Canadian Paint and Finishing.
    December 1974.
6.   "Electroless Electrocoat" now in Production  at Chrysler.  Industrial
    Finishing 51(10):72, October, 1975.
7.   Schrantz, Joe. How Autodeposited Coating Benefits Chrysler.   Industrial
    Finishing (51(11):14-22.   November 1975.
8.   Kachman, N.C.  Gener*1 Motors Corporation,  Warren,  MI.   Letter to
    G.M. Hansler, Region II,  Environmental Protection Agency,  New  York,
    N.Y.  February 11, 1974.
                                      95

-------
 9.   Conversation  of J.A.  McCarthy,  U.S.  Environmental Protection
     Agency with Chrysler  Corporation,  Detroit, MI.  November 12,
     1975.
10.   Boston Transportation Control  Plan,  Federal  Register,  p. 25166.
     June 1975.
11.   Coil Coaters  Discuss  Anti-Pollution  Systems.   Industrial Finishing.
     49(7):  48-54.   July  1973.
12.   Miller, E.P.  and D.D. Taft.   Fundamentals  of Powder  Coating.
     Society of Manufacturing Engineers.   Dearborn,  MI.   -1974.
13.   Powder Coating:  The  Problems  and  the Promises.   Products  Finishing.
     36(8), May 1972.
14.   High-productivity EPS System at Bunting.   Industrial Finishing.
     48(3), March 1972.
15.   Robinson, G.T.   Painting Stadium Seats.   Products Finishing.
     40(4):  50-54.   January 1976.
16.   Vinyl-Coated Wire Screens Resist Salt in  Air Filters Aboard Ship.
     Powder Coating World.  |J_(2):   26, 2nd Quarter 1975.
17.   Powder Coating Is a New Answer at  GTE Electric.   Powder Finishing
     World II_(2):   30, 2nd Quarter 1975.
18.   Powder System Highly  Satisfactory  at Rain Bird.   Powder Finishing
     World H(2):   33-34,  2nd Quarter 1975.
19.   High-Production Thin  Film Powder Coating  a Success  at Rival.
     Products Finishing.  39_(5):   32-87,  February 1975.
20.   Powder Coating Used as Insulation  for Magnet Wire.   Products
     Finishing.  39_(5):   94-95, February 1975.
21.   Miller, B.C.  and P.M. Yoder.  Hot  Melt Coatings.   In:  Industrial
                                       96

-------
     and Specialty Papers,  Volume  I  -  Technology.  Mosher, R.H. and
     D.  Davis (eds.).   New  York, Chemical  Publsihing Col, 1968.
22.   Reichner,  R.F.   Extrusion  Coating.   In:   1973-1974 Modern  Plastics
     Encyclopedia, New York,  McGraw-Hill,  Inc.,  1974.  P. 315-320.
23.   Roberts, A.G.  Organic Coatings,  Properties,  Selection  and Uses,
     U.S. Department of Commerce,  National  Bureau  of Standards, Washington,
     D.C.  February 1968.
24.   Electrostatic Spraying of  Water-Borne Paints.  Poll, G.H., Jr.  (ed.)
     Products Finishing 40_(4):   34-41, January 1976, p.  34-41.
25.   Berbeco, G.R. and S.V. Nablo.  Electron  Beam  Curing.  Paint  and
     Varnish Production.  64_(9):  39-42, August 1974.
26.   Hoffman, C.R.  Electron Beam Curing a Non-Polluting  System,  High
     Voltage Engineering Corporation,  Burlington,  Mass.
27.   Miranda, T.J. and T.F. Huemmer.  Radiation Curing of Coatings.
     Journal of Paint Technology.   41_(529):  118-128,  February  1969.
28.   Rybny, C.B.,et al.  Ultraviolet Radiation Cured Coating.   Journal
     of Paint Technology.   46J596):   60-69, September  1976.
29.   Billmeyer, F.W., Jr.   Textbook of Polymer Science.   Interscience
     Publishers of John Wiley and Sons, New York,  N.Y.  March  1966.
30.   Radiation Curing Goes Begging for Coaters.  Iron  Age.   p.  43-52,  August
     18, 1975.
31.   Shahidi, J.K. et al.   Multifunctional Monomers  for  UV  Cure.   Paint
     and Varnish Production.  August,  1974.
                                       97

-------
              4.0  COST OF VOLATILE ORGANIC CONTROL

4.1  Summary
      This chapter shows that many factors can affect the investment
and annual cost of hydrocarbon control  (Section 4.2.1);  thus  a  wide
range in control costs at existing facilities is anticipated.   Cost
estimates made for incineration and adsorption (4.2.2) indicate that
for low hydrocarbon concentrations (around 100 ppm), carbon adsorption
is the more economical of the two control  alternatives.   For control of
high hydrocarbon concentrations (around 25 percent of the LEL), carbon
adsorption is more economical and can even make money if recovered solvents
can be credited at market value; if no value is given for recovered solvents,
then incineration with primary heat recovery is more economical than
adsorption at high LEL concentrations.
      With regard to cost effectiveness, (cost per ton of volatile organic
removed), it is found that control of high concentrations is more economical
than is control of low concentrations by roughly one order of magnitude
(4.2.3).  It is beyond the scope of this chapter to evaluate the afford-
ability of various volatile organic control costs for the industries affected,
because each industry has a unique financial position.
4.2  General Discussion of Costs
      For most business ventures, an adequate return on investment is the
most important requirement to be evaluated before investment.  Air pollution
control often does not provide any return on captial investment—the objective
is to comply with a given emission standard.  The primary criteria for
selection is ability to achieve the required emission reduction.  If several
methods exist to control emissions, the problem then becomes one of
evaluating the cost of feasible alternatives to determine the preferred
alternative.
                               98

-------
      To evaluate the cost of feasible alternatives, it is important to
understand the factors that affect costs and the magnitude of these
costs.  Factors affecting the investment and annual  cost of control are
given in Section 4.2.1.  Costs of hydrocarbon control  for two types of
control technology is given in Section 4.2.3.  The discussion is aime'd
primarily at add-on control equipment.  It is beyond the scope of this
discussion to consider the cost of process and material changes to achieve
compliance with hydrocarbon regulations at existing facilities due to
the almost unlimited variety In processes and facilities.
4.2.1  Factors Affecting Investment and Annual Cost
4.2.1.1  Process Characteristics — Before one can estimate investment and
annual cost of control, one must first know certain characteristics (gas
stream volume and temperature, for example) of the source of pollution
because these characteristics affect selection and sizing of the control
equipment.  In some cases only one control alternative may be feasible;
however, in most cases two or more alternatives are feasible.  A good
knowledge of the specific source of pollution is advisable because process
or input material changes may be the least expensive method of controlling
volatile organic emissions.
4.2.1.2  Present Degree of Control -- Another important factor to consider
is the present degree of control of volatile organics versus the required
degree of control.  Industries may have control equipment but must achieve
more reduction in emissions to comply with a standard.  Upgrading or
modifying the existing equipment is usually the least expensive in these
cases but complete replacement of the existing control equipment with more
efficient equipment is another alternative.  Application of engineering
ingenuity is often needed to solve problems related to volatile organic
control at existing facilities.
                              99

-------
      If process modification or upgrading existing control  equipment is
not possible, the use of additional  add-on collection equipment must be
examined.   Once the types of controls capable of achieving the desired
control  level are chosen, one can choose the equipment with minimum total
installed cost.  (The cost of the control  equipment itself can sometimes
be a small fraction of the total system installed cost.)   Tables 4-1 and
4-2 show the items typically included in the investment and annualized
operating cost of add-on control systems.
4.2.1.3  Plant Facility Characteristics -- Availability of physical space
for the control system, utilities to operate the control  system, forms
and quantity of heat recovery possible at the facility, and ability to
utilize recovered products are classified as plant facility characteristics
that affect selection and costs of organic vapor control  technology.
4.2.2  Organic Vapor Control Costs
                                                                       »
4.2.2.1   Incineration -- Incineration can be an economical control alternative
if heat recovery techniques can be utilized.  To /illustrate the importance
of heat recovery, three cases were investigated:
      1.  No heat recovery.
      2.  35 percent primary heat recovery.
      3.  35 percent primary heat recovery and 55 percent secondary
          heat recovery of the remaining 65 percent.
      For each case, cost estimates  were made for three inlet flow rates
(5,000,  15,000, and 30,000 scfm), two inlet stream temperatures (70 and 300°F),
and three stream conc_,iirations (0, 15, and 25 percent of the LEL).  Other
assumptions used in developing the estimates are given in Table 4-3.
                              100

-------
      Table 4-1.   TYPICAL ITEMS INCLUDED  IN  INVESTMENT  COST
                          OF ADD-ON CONTROL  SYSTEMS
0 Basic Collection Equipment

0 Auxiliary Equipment

   >Air movement equipment

    Fans and blowers
    Hoods, ducts
    Electrical (motors, starters, wire conduits,  switches,  etc.)

   •Liquid movement equipment

    Pumps
    Electrical (motors, starters, wire conduits,  switches,  etc.)
    Piping and valves
    Settling tanks

   •Instrumentation for measurement and control  of:

    Air and/or liquid flow
    Natural gas and/or fuel oil flow
    Temperature and/or pressure
    Operation and capacity
    Power

0 Research and Development - this might include  gas  stream
  measurement, pilot plant operations, personnel  costs, etc.

0 Installation

    Labor to install
    Cleaning the site
    Yard and underground
    Building modification
    Inspection
    Support construction
    Protection of existing facilities
    Supervising and engineering
    Startups

0 Storage and Disposal Equipment

0 Contingencies

0 Sales Tax
                                101

-------
      Table 4-2.   TYPICAL ITEMS INCLUDED IN ANNUAL COSTS
                         OF ADD-ON CONTROL SYSTEMS
0 Capital Charges
0 Operating Costs
     Utilities needed to operate the control equipment
     Materials consumed (such as fuel) in operating the
     control system
     Waste disposal operations
0 Overhead
     Property taxes
     Insurance
0 Maintenance Costs
     Replacement of parts and equipment
     Supervision and engineering
     Repairs
     Lubrication
     Surface protection (such as cleaning and painting)
0 Offsetting Cost Benefits from Operating Control System
     (such as recovery of valuable by-product)
                                702

-------
             Table 4-3.  ASSUMPTIONS USED IN DEVELOPING COST
                         ESTIMATES FOR CATALYTIC AND NONCATALYTIC
                         INCINERATORS
0 Noncatalytic incinerators designed for both oil and natural gas operation.

0 Catalytic incinerators designed for natural gas and propane operation.

0 Catalytic incinerators capable of 800°F operation below 6 percent LEL;
  1200°F design capability for operation from 6 percent to 25 percent LEL.

0 3-year catalyst life.

0 Costs based on outdoor location.

0 Rooftop installation requiring structural steel.

0 Fuel cost of $1.50 million Btu (gross).  Correction factors are provided
  to determine operating costs at higher fuel prices.

0 Electricity at $0.03 kw-hr.

0 Depreciation and interest was taken as 16 percent of capital investment.
  Annual maintenance was assumed to be 5 percent of capital cost, taxes
  and insurance, 2 percent, and building overhead, 2 percent.

0 Direct labor assessed at 0.5 hr/shift x 730 shifts/yr x $8.00/hr =
  $2920/yr direct labor expense.

0 Operating time: 2 shifts/day x 8 hr/shift x 365 days/yr = 5840 hr/yr.
  Correction factors are provided to determine annual cost at different
  operating times.

0 The noncatalytic incinerator utilized was based on:

       * 1500°F capability.
       • 0.5-second residence time.
       • Nozzle mix burner capable of No. 2 thru No. 6 oil firing.
       •• Forced mixing of the burner products of combustion using a
         slotted cylinder mixing arrangement.  This cylinder allows
         the burner flame to establish itself before radial entry of
         the effluent thru slots in the far end of the cylinder.
       • A portion of the effluent to be incinerated is ducted to the
         burner to serve as combustion air.  This allows the burner
         to act as a raw gas burner, thus saving fuel over conventional
         nozzle mix burners.  This design can only be used, however,
         when the Op -~™tent of the ov»n exhaust is 17 percent by
         volume or above.

°The catalytic afterburner was costed for two design points, 800 and 1200°F.
the higher temperature design is required for LEL levels exceeding 6 percent.
(At 600°F into the catalyst and a 6 percent LEL, the outlet temperature of
the catalyst is approximately 800°F; at a 25 percent LEL condition and a
minimum initiation temperature of 500°F, the catalyst reaches an outlet
temperature of around 1200°F.
                             103

-------
      Based upon the results of the cost estimates,  cost curves  were
developed (Figures 4-1  through 4-15).   Because of the unique plant
facility characteristics, actual  control costs for some plants can be
substantially higher than estimates given here.   For figures that give  total
annual cost, cost effectiveness information is also presented, that is,
the cost per ton of hydrocarbon removed.  Cost effectiveness information
is a useful criterion when trying to devise air pollution control strategies
to reduce the total amount of a pollutant emitted at a minimum cost.
Cost effectiveness is discussed further in Section 4.-2.S.
Installed cost of incinerators — Figures 4-1, 4-6, and 4-11 give the installed
cost for incinerators designed for (1) no heat recovery, (2) primary heat
recovery, and (3) primary and secondary heat recovery.  The costs were intended
to represent typical retrofit situations.  However, further investigation has
revealed that the costs are more representative of the minimum retrofit
situation, essentially  the same as installation during the construction of
a new plant.  The installed cost in more typical retrofit situations will be
1.5 to 2 times the values shown in Figures 4-1, 4-6, and 4-11.  In very
difficult cases, the cost can be 3 to 5 times that shown in the figures.
      The average installed cost of incinerators with primary heat recovery
is roughly 25 to 30 percent greater than incinerators without heat recovery.
Incinerators with primary and secondary heat recovery have roughly 50 to 60
percent higher installed costs than incinerators without heat recovery.
Annual control cost of incinerators -- The annual control costs for  incinerators
given in this section include the  items shown in Table 4-3.   {Annual depreciation
is viewed in this study as a cost, not as a credit against taxable  income.)
Thus, the analysis is simplified,  but the annual control cost may be overstated.
However, given the wide range in cost of retrofitting control equipment at
existing facilities, this approach is deemed acceptable.)
                              104

-------
      Annual  control  cost curves are given in Figures  4-2 through  4-5,
Figures 4-7 through 4-10, and Figures 4-12 through 4-15 for the three
cases investigated.  The costs obtained from these figures include a fuel
cost of $1.50/million Btu and operating time of 5840 hr/yr.   If fuel cost
and/or operating time for a specific installation differ from these values,
the annual costs obtained from these figures can be adjusted by correction
factors obtained from Figures 4-16 through 4-19.  The annual costs read
from the figures are multiplied by the correction factor(s).  The  cost
effectiveness can be corrected by the following equation:
             CEc = CEi x Ff x Fh x 5840
                                   actual hours operatea

      where:  CE  = corrected cost effectiveness
              CE. = Cost effectiveness read from the appropriate figure
              F,.  = Correction factor for fuel cost
              F.   * Correction factor for hours operated
               h
      The costs given in annual cost figures include depreciation and interest
for the capital investment at a minimum retrofit cost situation.  In cases
where retrofit difficulties cause the installed cost to be increased
substantially, an appropriate multiplying factor (retrofit difficulty factor)
can be used to obtain the increased capital cost.  The increase in annual
cost, for a given retrofit difficulty factor, will  be a varying amount
for the different cases of vapor concentration, initial temperature, and
heat recovery.  The increase in the ann1"! cost for the different cases is
given in Tables 4-4 and 4-5.  The annual cost is first read from the
applicable figure, then is increased by the percentage given in Tables 4-4
and 4-5 under the appropriate difficulty factor.
                               105

-------
s

X
o
ee
ta
o
o
                 5        10        IS       20        25,       30


                             PROCESS FLOW, 103 scfm (APPROXIMATE)




           Figure 4.1. Capital cost for direct flame and catalytic afterburners without

           heat recovery (70 - 300 op process gas inlet) - Case 1.
                                   106

-------
 X

v>
o
o
     800
     700
     600
     500
     400
     300
    200
   • 100
                10         20         30



                         FLOW,scfmx1o3
40
50
                      7000



                      6000




                      5000




                      4000




                      3000




                      2000




                      1000
                  fe
                  o
                  o
                       400
                       300
                       200
                       100
                                                                                                 100
                                                                                                    ppm
                                                                                             J5 percent L EL
                                                                                            25 percent LEL
10        20        30



         FLOW,scfmx103
40
              Figure 4-2.  Annual cost and cost-effectiveness of direct flame incinerators.  (No heat recovery - process

              temperature = 70 °F) - Case 1.
50

-------
 X
V*
                                                                   7000
                10
 20        30

FLOW,scfmx103
40
50
10        20       30

         FLOW, scfmx103
40
50
               Figure 4-3. Annual cost and cost-effectiveness of direct flame incinerators (no heat recovery - process
               temperature -  300 °F) - Case 1.

-------
O
10
                       10
 20        30



FLOW, scfmx1()3
40
                                                                       I
                                           5000




                                           4000




                                           3000




                                           2000




                                           1000




                                              <




                                            400








                                            300








                                            200








                                            100
                                                                                                       tOOppm
                                                                                                   •Spwcent LEL
                                                                                                   ~~~—-




                                                                                                  25 percent LEL
10        20        30


         FLOW,sefmx103
                      Figure 4-4. Annual cost and cost-effectiveness of catalytic incinerators (no heat recovery - process

                      temperature = 70 °F) - Case 1.
                                                                                                                       40
                                                                                                SO

-------
—i  O
—J  —I
o  <
         700
         600
         500
         «»
         300
        ,200
         100
                     10        20         30
                              FLOW,scfmx103
40
50
                      4000
                      3000
                      2000

                                                                          1000 	
                                                                          400
                       300
                                                                           200
                       100
                                       15 percent LEL
                                      	-.
                                    25 percent LEL
10        20        30
         FLOW,scfmx103
                Figure 4-5.  Annual cost and cost effectiveness of catalytic incinerators (no heat recovery - process
                temperature = 300 °F) - Case 1.
                                                                                                                     40
                                                                  50

-------
    340
     320
     280
«   240
o
u
_1


t   200
a.
     160
    120
      80
                          10        li        20        25        30


                                   PROCESS FLOW, 1fl3 scfm (APPROXIMATE}
35
40
45
       Figure 4-6.  Capital cost for direct flame and catalytic afterburners with primary heat

       recovery (70 - 300 °F process gas inlet) - Case 2.
                                         m

-------
ro
           700
           600
           500
       •3,
       x   400
       <   300
           200
           100
                        10         20       30


                                FLOW,scfmx103
40
50
                      5000



                      4000


                      3000


                      2000


                      1000
                                                                              s
                  <   400
                  •*»
                                                                        V)
                                                                        o
                                                                        O
                       300
                       200
                       100
                                                                                                         100 ppm
10        20       30


         FLOW,scfmx103
40
                       Figure 4-7. Annual cost and cost-effectiveness of direct flame incinerators (primary heat recovery -
                       process temperature = 70 °F) - Case 2.
50

-------
co   3
     _!
     <
700
600
500
400
300
200
100
                                                             5
                                                             6"
                                                             O
                                                             o
                                                                         5000


                                                                         4000


                                                                         3000


                                                                         2000

                                                                         1000
                                                                 400
                                                                          300
                                                                          200
                                                                          100
                                                                                               lOOppm
            10        20         30

                      FLOW, scfmx1Q3
                                                   40
50
                                                                                          15 percent LEL


                                                                                      25 percent L EL
 20        30

FLOW,scfmx103
           Figure 4-8.  Annual cost and cost-effectiveness of direct flame incinerators (primary heat recovery -
           process temperature = 300 °F) - Case 2.

-------
 X
*»
u
-J
     600
     500
    400
    300
    200
;100,
                 10        20        30

                          FLOW,scfmx1o3
                                           40
50
                                                                 3000


                                                                 2000

                                                                 1000
                                                                  400
                                                              b*   30°
                                                              a
                                                              u
                                                                  200
                                                                      100
                                                                                                 15 percent LEL
                                                                                                      ^••—^^

                                                                                                      25 percent LEL
10        20        30

         FLOW,scfmx1o3
50
             Figure 4-9.  Annual cost and cost-effectiveness of catalytic incinerators (primary heat recovery -
             process temperature = 300 °F) - Case 2.

-------
e
»•*

x
o

_l
0
                 10
 20        30


FLOW, scfmxIO3
                                                         50
                                                                    3000




                                                                    2000



                                                                    1000
                                      &
                                      o
                                      u
                                                                      s
                                                                    400
                                                                     300
                                                                    200
                                                                     100
10        20        30


         FLOW,scfmx103
                                                                                               SO
               Figure 4-10. Annual cost and cost effectiveness of catalytic incinerators (primary heat recovery

               process temperature = 70 °F) - Case 2.

-------
    400
    360
    320
    280
    240
<   200
    160
    120
                         10        15       20       25       30

                               PROCESS FLOW, 1fl3 scfm (APPROXIMATE)
35
40
45
 Figure 4-11.  Capital cost for direct flame and catalytic afterburners with primary and second-
 ary heat recovery (70 - 300 °F process gas inlet) - Case 3.

-------
o
u
    200
    100
                10
 20        30


FLOW,scfmx103
50
                                                                o
                                                                u
                                                                    5000



                                                                    4000



                                                                    3000



                                                                    2000



                                                                    1000
                                                                     400
                                                                     300
                                           200
                                           100
                                                                              100 pp
                                                                                                  J5 percent LEL



                                                                                                 25 percent LEL
 20        30


FLOW,scfmx103
                        Figure 4-12. Annual cost and cost-effectiveness of direct flame incinerators (primary and

                       secondary heat recovery - process temperature = 70 °F) - Case 3.

-------
00
              700
              600
              500
    400
         o
         u
<    300
    200
     100
               [0      _.  10        20         30


               	            FLOW.scfmx103
                                               40
50
                                                                    5000	




                                                                    4000 —




                                                                    3000 —




                                                                    2000 —



                                                                    1000 —
                                                                               s
                                                                          <   400
                                                                 ft
                                                                 o
                                                                 o
                                                                              300
                                                                              200
                                                                              100
                                                                                                          lOOppm
                                                                                                       15 percent LEL
10        20        30


         FLOW.scfmx103
40
                      ..  Figure 4-13. Annual cost and cost-effectiveness of direct flame incinerators (primary and secondary

                         heat recovery - process temperature = 300 °F) - Case 3.
50

-------
500
     400
w
o
v*
 X


I-*

o   300
o
_i
<
200
100
            10         20         30


                      FLOW, scfm x 1fl3
                                                          10
                                                                      3000


                                                                      2000


                                                                      1000
                                                                   s
                                                                  400
                                                             I

                                                             §   300
                                                             O
                                                                       200
                                                                       100
ID        20        30


          FLOW, scfm x103
            Figure 4-14.  Annual cost and cost-effectiveness of catalytic incinerators (primary and secondary
            heat recovery - process temperature = 70 °F) - Case 3.
                                                                                                                  40
SO

-------
   600
   500
o  400

X
   300
  200
   100
               10         20        30


                         FLOW,scfmx1o3
40
50
                                                                   o
                                                                   o
                        3000


                        2000


                        1000



                          s

                         400






                         300






                         200






                         100
                                                                                                  lOOppm
                                                                                               ,15 percent LEL
10         20         30


          FLOW,scfmx1<)3
                   Figure 4-15. Annual cost and cost-effectiveness of catalytic incinerators (primary and secondary

                   heat recovery - process temperature = 300 °F) - Case 3.
40
50

-------
                        THERMAL INCINERATION   Ti = 300°F
ro
                        NO HEAT RECOVERY     **

                        PRIMARY H.R.

                        PRIM & SEC H.R.
                                  3.0             4.0

                             FUEL COST, S/MM BTU
         THERMAL INCINERATION  Ti = 70°F
\     2.0
 s
  \
   \
    \
     \
      \
      \
       \
    3.0             4.0

FUEL COST. $/MM BTU
                         Figure 4-16. Factors to correct annual cost of thermal incineration for varying fuel cost.

-------
                       CATALYTIC INCINERATION  Ti = 70*F
                                                                                CATALYTIC INCINERATION  Ti = 300°F
ro
ro
      oc
      o
o
UJ
oc
cc
o
o
                                  3.0             4.0





                           FUEL COST.($/MM BTU)
         0.7
                                                                                   ^
                                                                                           3.0             4.0





                                                                                          FUEL COST, $/MM BTU.
                                                                                             6>
                        Figure 4-17. Factors to correct annual cost of catalytic incineration for varying fuel cost.

-------
                        THERMAL INCINERATION Ti = 70°F
                                                           THERMAL INCINERATION  Ti = 300°F
(Si
          1.5
       g 1-25
       o
       LU
       oc
       K
       O
       CJ

       v>
       o
       o
         0.75
          0.5
                       .NO HEAT RECOVERY
                        PRIMARY H.R.
                        PRIM & SEC H.R.
           2080
      5840

OPERATING TIME, hours/year
8760
                                                                                                 I
2080
      5840

OPERATING TIME, hours/year
8760
                      Figure 4-18.  Factors to correct annual cost of thermal incineration for varying operating time.

-------
IV)
-p.
            1.5
                          CATALYTIC INCINERATION  Ti = 70°F
                                                                                CATALYTIC INCINERATION Ti = 300°F
            1.25
oc
cc
o
CJ
fe
            1.0
           0.75
            0.5 -
                          PRIM
                             I


                    HEAT RECOVERY
                      ARY H.R.
                      &SECH.R.
                          NO
                          PRIMARY
             2080
                           5840
                      OPERATING TIME, hours/year
8760
                                                                                                     NOTE: 25% L EL WITH PRIM H.R
                                                                                                      OR PRIM & SEC H.R. EXCEEDS
                                                                                                       DESIGN TEMPERATURE OF
                                                                                                            INCINERATOR
                                                                            2080
     5840
OPERATING TIME, hours/year
                                                                                                                 8760
                        Figure 4-19.  Factors to correct annual cost of catalytic incineration for varying operating time.

-------
        Table 4-4  INCREASE IN ANNUAL COST OF DIRECT FLAME INCINERATORS
                         DUE TO RETROFIT DIFFICULTY FACTORS
   Vapor         Process      Heat
Concentration  Temperature  Recovery
Percent Increase in Annual  Cost
 At Retrofit Difficulty Factor:
   1.5          2          3
100 ppm
15 percent LEL
25 percent LEL
100 ppm
15 percent LEL
25 percent LEL
100 ppm
15 percent LEL
25 percent LEL
100 ppm
15 percent LEL
25 percent LEL
100 ppm
• 15 percent LEL
25 percent LEL
100 ppm
15 percent LEL
70°
70°
70°
300°
300°
300°
70°
70°
70°
300°
300°
300°
70°
70°
70°
300°
300°
None
None
None
None
None
None
Primary
Primary
Primary
Primary
Primary
Primary
Pri . & Sec.
Pri. & Sec.
Pri. & Sec.
Pri. & Sec.
Pri. & Sec.
2
4
5
3
4
7
4
8
17
5
10
20
8
16
80
10
25
5
7
10
6
8
15
9
16
33
10
20
40
16
32
160
20
50
9
14
20
11
16
30
18
32
66
20
40
80
32
64
320
40
100
                              125

-------
Table 4-5  INCREASE IN ANNUAL COST OF CATALYTIC INCINERATORS
               DUE TO RETROFIT DIFFICULTY FACTORS
                              Percent Increase in Annual Cost
Vapor Process Heat At Retrofit Difficulty Factor:
Concentration Temperature Recovery 1.5 2 3
100 ppm
15 percent
25 percent
100 opm
15 percent
25 percent
100 ppm
15 percent
25 percent
100 ppm
15 percent
25 percent
100 ppm
15 percent
25 percent
100 ppm
15 percent

LEL
LEL

LEL
LEL

LEL
LEL

LEL
LEL

LEL
LEL

LEL
70°
70°
70°
300°
300°
300°
70°
70°
70°
300°
300°
300°
70°
70°
70°
300°
300°
None
None
None
None
None
None
Primary
Primary
Primary
Primary
Primary
Primary
Pri. & Sec.
Pri. & Sec.
Pri. & Sec.
Pri. & Sec.
Pri . & Sec .
6
6
8
8
9
11
6
11
22
10
15
16
12
18
36
14
25
13
13
15
16
17
22
13
22
44
20
30
33
24
36
72
28
50
25
26
30
33
34
43
25
45
88
40
60
65
48
72
144
56
100
                      126

-------
     Three observations regarding catalytic incinerators  are offered  for
those choosing between a catalytic and noncatalytic incinerator.   First,
a catalytic incinerator will probably oxidize any sulfur  in the fuel  to
SOo which will subsequently form sulfuric acid mist.   Therefore,  it is
recommended that only fuels that are essentially sulfur free, i.e.,
natural gas and propane, be used with catalytic incinerators. Second,
the catalytic element may be blinded by any particulate matter in the
exhaust gases to be incinerated; the blinding reduces effectiveness
because the catalyst contact area has been decreased.  Third, catalysts
cannot operate above approximately 1200°F and still give  reasonable
service life  (2 to 3 years).  In general, catalyst manufacturers will
limit their applications to streams with concentrations no higher than
25 percent of the LEL.

     Primary heat recovery (oreheat of the effluent from  the oven before
it enters the incinerator) reduces the fuel rate in the incinerator
significantly regardless of inlet temperature or LEL condition.  However,
at the higher LEL concentrations, the requisite heat exchanger presents
problems for both catalytic and noncatalytic incinerators.  At ^he
assumed 35 percent efficiency of the heat exchanger, the  design limita-
tions for the incinerators may be exceeded (for example,  the catalytic
unit may be forced to operate at temperatures greater than 1200°F, which
would shorten the life of the catalyst).  This problem can be solved by
using a bypass around the primary heat exchanger, in effect creating
a heat recovery unit of variable efficiency.
     Annual control cost curves for incinerators with primary heat
recovery are presented in Figures 4-7 and 4-10.

                            127

-------
     An imnortant assumption used in developing the cost estimates for
Case 3 is that all of the recovered secondary heat is utilized.
Section 3.2.2 discusses the implications of this assumption.
4,.2.2.2  Adsorption  -- Whereas incineration relies on heat recovery techniques
to offset annual operating cost, adsorption systems rely primarily on
solvent recovery and/or heat recovery to offset annual operating cost.
Many different configurations could be costed out.  To simplify the
analysis and presentation of results, this section is limited to carbon
adsorptionsystems with solvent recovery,  the system that is more commonly
used.
     Several assumotions were made in developing the cost estimates presented
in this section.  These basic assumptions are listed in Table 4-8.  In
addition to these assumptions, several  assumptions were made when defining
the control  system before the cost analyses were initiated.   The approach
taken was to use Reference 2 as a source for such design information as horse-
power requirements and water requirements for the various sized control systems.
Where possible, assumptions for adsorption systems were chosen to be identical
to assumptions for incinerators (see Table 4-3) to provide a direct cost
comparison between adsorption systems and incinerator systems.  The recovered
solvent was  assumed to be valued at (1) zero value, (2) $1.50 per million
Btu based on using the recovered solvent as a fuel, (3) $0.85 per gallon fc"
benzene and  $0.465 per gallon for hexane based on selling the recovered solvent
at its current market value.
      The assumptions do not provide for any distillation equipment or water
treatment facilities.  If any water soluble compounds are encountered in the
vapor, the costs will be increased considerably.  The costs do not include
                             128

-------
  Table 4-8.   ASSUMPTIONS USED IN DEVELOPING COST ESTIMATES  FOR
                              CARBON ADSORBERS
0 Exhaust gases contain benzene and hexane (50/50 weight percent)
  mixture in air.
0 Exhaust gas temperatures of 70,  170,  and 375°F.
0 Hydrocarbon concentrations of 100 ppm,  15 percent of the LEL  and
  25 percent of the LEL.
0 Exhaust gas flow rates of 1,000, 10,000, 50,000 scfm.
0 Fuel costs of $1.50/million Btu.
0 Electricity at $0.03/kw-hr.
0 Activated carbon at $0.68/lb.
0 Water at $0.04/thousand gallons.
0 Steam at $2/thousand Ib.
0 5-year life of activated carbon.
0 Adsorber operating at 100°F.
0 Market value (December 1975) of benzene = $0.85/gallon;
  market value (December 1975) of hexane  = $0.465/gallon.
0 Normal retrofit situation.
0 Direct labor assessed at 0.5 hr/shift x 730 shifts/yr x $8/hour =
  $2920/yr.
0 Annual maintenance, taxes, insurance, building overhead, depreciation,
  and interest on borrowed money taken  as 25 percent of capital investment.
0 Operating time = 5840 hr/yr.
                                129

-------
any particulate removal  equipment.   Compounds  which  are  difficult  to  adsorb
or desorb, if present, will  also add considerably to the installed and
operating costs.
Installed cost of adsorbers  -- Capital  cost for adsorption systems designed
to recover the solvent is given in Figure 4-20   Based upon Figure 4-20,
the unit capital costs for two gas flow rates  are given  in Table 4-9.
      For a given gas flow rate in Table 4-9,  the higher unit costs are
for adsorption of gases at 25 percent of the LEL and the lower unit costs
are for concentrations of 100 ppm.
                 Table 4-9.  UNIT CAPITAL COST FOR CARBON
                                  ADSORPTION SYSTEMS
              Gas flow rate,                    Control  cost,
                  scfm                            $/scfm
                  5,000                           30 - 36
                 30,000                           18 - 21
     Annual control cost for adsorbers -- Figures 4-21  through 4-23 are
given to illustrate the annual costs of carbon adsorption systems over a
representative size range.  These figures and Table 4-10 illustrate the
importance of the value of the recovered solvents.
                               130

-------
                                                                              O
                                                                              u
                                                                              Q)
                                                                              o
                                                                              o
                                                                              V)
                                                                              -a
                                                                              CO

                                                                              •a
                                                                              CD
                                                                        CJ

                                                                        K
                                                                              LLJ

                                                                              CJ
                                                                              
-------
co
ro

 x

 CO
 o
 u
 _l
 <
 =3
 z
 <
              350
             300
             250
             200
             150
             100
              50
A -100 ppm; 70° EXHAUST TEMPERATURE
B  15 percent LEL; 170°F EXHAUST TEMPERATURE
C  15 percent LEL, 375°F EXHAUST TEMPERATURE
D - 25 percent LEL;170°F EXHAUST TEMPERATURE
E  25 percent LEL; 375°F EXHAUST TEMPERATURE
         10        20        30

           ADSORBER CAPACITY, scfm x
                                                       40
                                                        50
                                                                   6000

                                                                   5000


                                                                   4000


                                                                   3000

                                                                   2000


                                                                   1000
                                                             s
                                                            300
                                                            200
                                                            100
10        20        30        40

  ADSORBER CAPACITY, scfm x 1fl3
                                                                                                                        50
                          Figure 4-21. Annual cost and cost-effectiveness of carbon adsorption systems (no credit given for
                          recovered solvents).

-------
CO
co
            o
            x

                400
350
300
                250
                200
                150
100
 50
                        A 100 ppm; 70 °F EXHAUST TEMPERATURE
                        B 15 percent LEL; 170 °F EXHAUST TEMPERATURE
                        C • 15 percent LEL; 375 °F EXHAUST TEMPERATURE
                        D 25 percent LEL; 170 °F EXHAUST TEMPERATURE
                        E 25 percent LEL; 375 °F EXHAUST TEMPERATURE
                            10        20        30        40
                               ADSORBER CAPACITY, scfm x 103
                                                    50
                                                                7000

                                                                6000

                                                                SOOO

                                                                4000

                                                                3000

                                                                2000

                                                                1000
                                                                             o   400
                                                                                 300
                                                                                 200
                                                                                 100
10        20        30
   ADSORBER CAPACITY, scfm x
                                                                                                           40
50
                           Figure 4-22. Annual cost and cost-effectiveness of carbon adsorption systems (recovered solvent
                           credited at fuel value).

-------
co
            300
            200
            100
X
Vi
V3

O   -100
U
            200
            -300
            -400
            -500
                        10        20        30
                                 FLOW,scfmx103
40
                                                                      o
                                                                      •*;
                                                                      O
                                                                      U
                                                                          300
                                                                          200
                      100
                     -100
                                                                          200
                             LEGEND:
                               A 100 ppm; 70 °F EXHAUST TEMPERATURE
                               8  1i % LEL; 170 °F EXHAUST TEMPERATURE
                               C  1S % LIL; 375 °F EXHAUST TEMPERATURE
                               D  2S % LEL; 170 °F EXHAUST TEMPERATURE
                               E  25 % LEL; 375 °F EXHAUST TEMPERATURE
                                                                              10        20        30
                                                                                       FLOW,scfmx103
                                                                                                                  40
                                                                        50
                      Figure 4-23. Annual cost and cost-effectiveness of carbon adsorption systems (recovered solvent
                      credited at market chemical value).

-------
          Table 4-10.   ANNUAL COST FOR CARBON  ADSORPTION  SYSTEMS

Value

No value
Value = fuel value

of recovered
solvent


Value = market chemical value
Annual cost, 10 $

5,000
47a-
35b-
(15)b

scfm
65b
47a
- 40a

30,000
145a -
65b-
(255)b

scfm
230b
140a
- 130a
Low concentrations.
High concentrations; parentheses indicate a net profit.
4.2.2.3  Cost Comparisons,  Incineration Versus  Adsorption  ~   Based  upon
the cost analysis in this section,  the following conclusions  are  indicated:
      1.  For control  of low hydrocarbon concentrations  (approximately
          100 ppm), carbon  adsorption is more economical than incineration
          if there are no water soluble compounds in  the vapor, no compounds
          which are difficult to adsorb or desorb,  and no  particulate
          matter.
      2.  For control  of high hydrocarbon concentrations  (approximately
          25 percent of the LEL), carbon adsorption is more economical
          than incineration only if recovered solvent can  be  credited
          at market value and none  of the above-mentioned  types of material
          are encountered.
      3.  Incineration with primary heat recovery is  more  economical than
          adsorption at high concentrations if  no value  is given  for recovered
          solvents.
                                 135

-------
4.3  REFERENCES FOR CHAPTER 4

     1.  Interim report on cost of incineration.   Combustion  Engineering,
         Air Preheater Division, Wellsville,  New  York.   Prepared for
         U.S. Environmental Protection Agency,  Research  Triangle Park, N.C.,
         under Contract No. 68-02-1473.   December 1975.

     2.  Hydrocarbon Pollutant Systems Study.   Vol.  II.   MSA  Research Corp.
         Prepared for U.S. Environmental  Protection  Agency, Research Triangle
         Park, N.C., under Contract No.  EHSD  71-12.   Publication No. APTD-1500.
         January 1973.
                                    136

-------
                      5.0  APPROACHES TO DETERMINATION
                      OF TOTAL "NONMETHANE" HYDROCARBONS
5.1  Summary
     No single, practical emission measurement method currently exists
that can be generally used to determine total nonmethane concentrations in
all situations.  If used selectively, however, several hydrocarbon sampling
and analytical techniques that yield acceptable results are available for
specific applications.  This section briefly explains each of these techniques
and provides application guidelines.  Supplementary measurements needed to
determine an emission rate are also suggested.  For this section "hydrocarbons"
is used interchangeably with "volatile organics".
15.2  Introduction
     Where applicable, a material balance is the most accurate measurement
technique available.  The technique relies on the fact that what goes into
the process must exit the process, unless chemical change occurs.  It is
useful  for such sources as spray booths and low temperature ovens where control
is by carbon adsorption.  The solvent content for coatings can easily be
determined and the amount of solvent recovered can be accurately measured.
For chemically reactive coatings, e.g. two component coatings, correction
for the reaction products is necessary.  Some problems can occur if separation
of evaporation quantities is necessary, e.g., for spray booths and following
ovens.   In such cases it would probably be necessary to measure the emission
from at least one of the exhaust streams to determine the proper apportion-
ment of emissions.
                                     137

-------
     For those cases where a material  balance can  be  made,  the  recommended
approach is for the sources to be required  to keep records  of solvent
used (including diluent and cleaning solvent) and  solvent  recovered.   A  one
week check by the control  agency could be used to  verify the plants  record-
keeping system.
      Source  evaluation  by  emission  measurement can be applied when it is
 not possible to  generate  emission values of  sufficient accuracy through a
 material  balance,  or when  stack  gases  have been oxidized or otherwise affected
 to  an  unknown extent.   Emission  measurement  methodology involves two separate
 but related  steps:   sample collection  and sample  analysis.  Before using
 any specific test  method,  however,  the question of vhether the collection and
 analytical techniques will  satisfy  the true  intent of the hydrocarbon emission
 standard  must be resolved.
      Although the  definition  of  any pollutant in  a desired or proposed  emission
 standard  is  normally the  point of reference  for designing the emission  measurement
 method, it must  be recognized that  the measurement method used can modify
 the pollutant Mdefinition."  Therefore,  to satisfy the true intent of the
 emission  standard, attention  must be  given to the proper selection of
 measurement  methodology and to the  elimination of vagueness in the specifications
 of  the selected  method.   Consider for a  moment a  definition of nonmethane
 hydrocarbons as  "those  hydrocarbons,  methane excepted, which exist in the
 gaseous state or behave as a  gas at the  point of  measurement."  The words
 "behave as a gas"  are  important  from  a collection standpoint, because it
 may not be possible to  separate  the gaseous  from  the particulate hydrocarbons
 without inducing a mechanism  that shifts the original gas/particulate ratio.
                                     138

-------
For example, if a sample collection method specifies  a  glass  wood  plug  or
similar filter in the gas sample probe  to prevent  participate  hydrocarbons
from reaching the sample container, gaseous hydrocarbons  might  adsorb on
the filter, or they might condense on particulate hydrocarbons  or  other
particulate condensation nuclei.  Particulate hydrocarbons  initially entrapped
on the filter could, on the other hand,  vaporize during the course of sampling.
When sampling occurs, where acids are present,  such as  at an incinerator out-
let, the filter could serve as a site for reactions between acids  and hydro-
carbons.  Moreover, by not specifying that the  sample probe is  to  be heated
to stack temperature, an additional question could  be raised as to the
possible loss of sample resulting from condensation in  the probe.
     Analytical requirements should also be indicated in the definition of
nonmethane hydrocarbons.  For example, a definition that states "...measured
on a carbons basis..." specifies that all the carbon atoms present in the
sample that result from nonmethane hydrocarbons are to  be counted.  Omission
of this clause would imply that the analysis method must report the entire
mass of the nonmethane hydrocarbon compounds.
    . Because of variations in the makeup of the hydrocarbons emitted, each
process tested must be individually evaluated to ensure that the collection
and analytical methods employed are consistent with the definition of non-
methane hydrocarbons.
5.3  Measurement Approaches
     General requirement for obtaining valid air pollution measurements are
covered quite well in the literature  '  and will not be restated here.   The
currently employed hydrocarbon source measurement techniques are limited
to the "extractive" classification, which means the gas must be withdrawn

                                     139

-------
from the source before it can be analyzed.
     Normally, process off-gas is extracted at the centroid of the  stack
or duct, except when there is reason  to suspect that the hydrocarbon  con-
centration varies within the stack cross section.   In that event,  the  sample
must be obtained by a procedure known as traversing, that is the sample
must be obtained from the centroids of a number of equal  areas within  the
cross section.  (The number selected is in  direct proportion to the expected
degree of irregularity in the exhaust gas.)
     Variations in gas concentrations are most often encountered where
relatively quiescent zones exist in the stack cross section.  These zones
are usually revealed by a velocity traverse.  If several  ducts that serve
different processes or vent noncontinuous operations are  manifolded together,
however, insufficient gas mixing could result in unpredictable concentration
variations — the presence of which would not be disclosed by a velocity
traverse.  Either of the above situations are best revealed by traversing
the cross section with a continuous-response hydrocarbon  detector.   In
addition, persons conducting the test must  be familiar with the processes served
by the stack to assure representative sampling of the emissions.
     Depending on the requirements of the emission standard, either a  grab
(instantaneous) sample or an integrated sample (a sample  collected at  a rate
proportional to the stack flow rate for a specified time  period) may be taken-
Three consecutive samples are normally collected for each test.
     If the mass emission rate is to be measured, simultaneous or near simul-
taneous measurements of stack flow rate are required.  Unless the stack gas
mixture is known (air, for example), its molecular weight must be determined.
As most analytical techniques are on a wet  basis, the water content of the
                                     140

-------
exhaust must also be measured to correct flow rates  and hydrocarbon  concen-
trations to standard conditions.  EPA Methods 2,3,  and  4,  respectively,  will
normally suffice.
     If the emission concentration varies and the emission standard  is a
time-averaged value, grab samples may be taken in sufficient quantity and
spread over a given time period to yield an average  value.  When  the stack
flow rate also varies, it will have to be measured  at the  time each  grab
sample is taken in order to proportion the results.   Grab  samples may be
analyzed directly, or be held for subsequent recovery and  analysis.   In-
tegrated gas samples may be collected without alteration,  or the  hydrocarbons
can be concentrated with a sorbent.  If the emission concentration is suspected
to vary erratically with time, it is best to use the integrated sample approach
thus reducing the total number of samples to be analyzed.   In some situations.
a continuous hydrocarbon detector can be integrated into the sampling system
with its data averaged either manually or automatically over a given time
period.
     The next two portions of this section are devoted to  explanations and
discussions of sampling and analytical techniques  that show promise  for  the
determination of total nonmethane hydrocarbons.  These techniques can, except
as noted, be combined in a variety of ways to yield an emission measurement.
The best combination for a particular source category will depend on the
intent of the regulation or emission standard in question.
5.4    Sanpling
     General sampling techniques and precautions have been extensively
         235
reported; ' »  those most suited to collection of total nonmethane hydrocarbons
are discussed below.
                                    141

-------
Activated Carbon Adsorption -- First investigated  for use  as  an  air
sampling method by Kupel  et al.,   an activated  carbon collection device
(Figure 5-1) was further  examined by Mueller and Miller in 1974.
Figure 5-1 shows a typical  sample device.   Once adsorbed on the
activated carbon, hydrocarbons are subsequently desorbed with a
solvent—usually carbon disulfide.  The hydrocarbon sample, now  a
liquid, is then analyzed  by a detector that is  insensitive to the
solvent.
      Advantages of the carbon adsorption tube  are:
      1.  Small sample quantities are collected.
      2.  Leaks are less  likely to occur, in comparison with bag samples.
      3.  Analysis of each sample can be repeated  several  times
          because only a  few milliters of the solution are
          required.
      Disadvantages include:
      1.  Gas temperature must be below 125°F for  adsorption to  be
          effective; precoolers are often required.
      2.  Moisture in the stack gas may condense and plug  the carbon
          bed.
      3.  A total gas sample is not collected,  so  hydrocarbon "break-
          through" must be prevented.  Because  a second carbon section
          is needed in each tube to demonstrate collection efficiency,
          each sample becomes two samples, thus doubling the number  of
          analyses required.
      4.  Collection efficiency for noncondensible and some volatile
          organics is poor.7
      5.  Sample desorption efficiency is poor  foe some volatile
          organics, particularly polar compounds.
      6.  Organic compounds could be displaced  from carbon by more
          readily adsorbed gases.7
      7.  The technique is not adaptable to total  combustion analysis.
                              142

-------
                                   BACKUP SECTION
                                FOR TEST VALIDATION,
                                    50 mgCARBON
 MAIN SECTION
FOR ANALYSES,
100 mgCARBON
oo
                                                                                                \
                                                                                                y
                                              AIR FLOW
                                    Figure 5-1.  Carbon-filled tube employed to trap solvent vapors.?

-------
                                                                       o
 Universal  Collector -- The  Universal Collector  described by Isbell

 also employs activated carbon,  but only to collect the more volatile

 organic  compounds (see Figure 5-2).   A preadsorber of Tenax GS  is  used

 to collect hydrocarbons that are difficult to desorb from activated

 carbon.
SAMPLE
 INLET
                   OUTLET
                 HEAVY
                  ENDS
                 OUTLET
       CARRIER
        GAS
        INLET
 LIGHT
 ENDS
OUTLET
          CLOSED
                              QUICK-
                             CONNECT
                             FITTINGS
  u
  o
  X
o
cj
cc
                      u
                      GJ
CJ
a
X
et
  O
  o
  cc
                                       u
                                       u
    SAMPLE COLLECTION                          SAMPLE RECOVERY

         Figure 5-2. Schematic diagram of universal collector.^
                               144

-------
     The adsorbers are maintained at or below ambient temperature  during
sample collection.  Recovery is obtained by raising the temperature to
approximately 200°C and flushing the adsorbers with a carrier gas.
     Advantages of the Universal Collector over activated carbon alone  are:
     1.  Sample collection and recovery efficiencies are improved.
     2.  The method is adaptable to total  combustion analysis.
     3.  Sample recovery procedures are simplified.
     Disadvantages include:
     1.  Two collectors in series must be used to demonstrate good collection
         efficiency.
     2.  Operating temperature during sampling affects collection  efficiency.
         Efficiency is best at low temperature but if the temperature is  too
         low, water condensation may plug the device.
     3.  Because extensive collection and recovery efficiency data are
         lacking, source components should be checked before the procedure
         is used.
     4.  Only one analysis can be made of each sample.
Cold Trap/Evacuated Cylinder - Condensable organics can be concentrated with
a cold trap, which normally consists of a piece of stainless steel  tubing
immersed in a refrigerant.  Midget impingers, however, both with and without
tips, can be used in  place of the tubing.  A pump or an evacuated  cylinder
can be used to move the sample gas, if necessary.  If a pump is used, the
cylinder must be teed-in between the second cold trap and the pump.
Collection occurs When the pollutant vapor is sufficiently cooled  to condense
and deposit on the interior surfaces of the trap.  Figure 5-3 is a generalized
                                                           g
schematic of the sample system described by Gadomski et al.
                                    145

-------
STACK
                     PROBE
                  HEATING TAPE
                                                        METERING VALVE
                                                                  EVACUATED
                                                                   CYLINDER
                                       COLD TRAP
  t
       DIRECTION OF FLOW
                                                      VACUUM GAUGE
0
Figure 5-3. Schematic of cold trap/evacuated cylinder sampling.
                                         146

-------
Six inches of glass wool in the tubing immediately preceeding  the  metering

valve prevents condensation carryover into the evacuated cylinder.   Both

the trap and the cylinder are analyzed for hydrocarbons.

     Advantages of the sampling system are:

     1.  Collection efficiency should approach 100 percent because the
         device is a total  gas sampler.   (A total  sample is not collected
         if a pump is used; collection efficiency  must then be demonstrated
         by using a second cold trap placed in series  with the first.)
     2.  If probe heating is not required, no electrical devices
         are required at the stack sample site.
     3.  No restrictions are made on the analytical  technique(s) used.


     Disadvantages include:

     1.  If condensable organics are carried into  the  cylinder, recovery
         may be incomplete.
     2.  Water in the stack gas may cause the cold trap to freeze.  This
         difficulty may be avoided if the gas first passes through an
         impinger with the tip removed.
     3.  Only one analysis per sample can be made.
     4.  Proportional sampling is difficult if the cylinder vacuum is the
         sole stack gas withdrawal force.
     5.  Immersion of traps into refrigerant prior to  sampling is
         accompanied by a lowering of internal pressure and a  corresponding
         influx og gas must be taken into account  when calculating sample
         volume.

Syringe or Purge Flask - Relatively simplified grab samples can be collected

with either glass syringes or purge flasks.  If condensation problems are

expected, either may be preheated with heat tape and maintained at an elevated

temperature during the short sampling period with  an insulated casing.

Although the syringe is obviously a manual technique,  the purge flask can  also

be manually operated with a rubber, one-way squeeze bulb.  After it has been

repeatedly flushed with stack gas, the syringe or purge flask  is allowed

to equilibrate to stack pressure before it is sealed.   The syringe is difficult

to leak check, but the purge flask is not,  If the volume of the purge flask is


                                     147

-------
large relative to the quantity of gas required for analysis,  several
analyses may be performed on the same flask before the pressure  is
appreciably decreased.  On the other hand,  the syringe is  normally
exhausted by one analysis.  If sample adsorption on the walls of either
device is appreciable, sample recovery may  be enhanced by  using  a heat
tape to raise the temperature.
     The chief advantages of both techniques are:
     1.  They are least expensive of hydrocarbon sampling  techniques; no
         reagents are required.
     2.  Electrical  devices are not required at the sampling  site,  which
         is a safety consideration at sites where the atmosphere may be
         explosive.
     3.  Collected samples are suitable for various analytical techniques.
     Disadvantages include:
     1.  It is difficult to detect leaks and perform leak  checks on a
         syringe.
     2.  Multiple samples are required for  time-averaged emission
         determinations.
     3.  It may be difficult to obtain repeatable analyses with  identical
         samples unless sample recovery techniques are automated.
Collapsible Plastic Bags - Collapsible plastic bags have been used  to collect
integrated gas samples for several years.  EPA Method 3 employs  a flexible
bag for samples to be analyzed for carbon dioxide, excess  air, and  dry
molecular weight.   More recently, EPA has  suggested that  Tedlar^bags  be
used to collect samples for measurement of  vinyl chloride  emissions.
     The sampling train (see Figure 5-4) is intended to be used  on  stack
exhausts having a relatively low moisture content, because water condensation
                                      148

-------
vo
           GLASS WOOL FILTER
           REVERSE "S"-TYPE
              PITOT TUBE
                                                  TEFLON
                                                SAMPLE LINE
                                                                       VACUUM LINE    NEEDLE VALVE
                      STACK WALL
FLOW METER
                                                                                 RIGID, LEAK-PROOF
                                                                                    CONTAINER
                                                Figure 5-4. Integrated bag sampling train.11

-------
could affect the hydrocarbon concentration.   If water condensation  is  a  pro-
blem, prediction of the stack gas at a constant rate with  clean  dry  air
often has been suggested as a remedy.  The bag must be subjected  to a
rigorous leak check before use.
     Tedlar^has been shown to be a superior bag material,    but  aluminized
Mylar* may work equally well.  Generally, any bag sample should be  analyzed
within 24 hours to minimize adsorption losses.    Other techniques  for
minimizing losses would include increasing the bag size (maximizing the  volume/
area ratio) and preconditioning the bag with sample gas.
     Advantages of bag samples are:
     1.   Because a total air sample is provided, collection efficiency
         should approach 100 percent.
     2.   Sample recovery requires no intermediate procedure; the  line
         is merely connected to the analyzer.
     3.   The quantity of each sample permits many reanalyses.
     4.   The quantity of sample is sufficient to-i|erve other purposes,
         such as odor threshhold determinations.
     Disadvantages include:
     1.   The bag cannot be readily heated to enhance sample recovery.
     2.   Higher molecu^gr weight compounds have a tendency  to adsorb
         on bag walls.
     3.   The size of the sample and the rigid container may create  problems.
         As with  any sample, it should be protected from sunlight.
     4.   Bags have a greater tendency to develop leaks than any other
         sampling device; frequent leak checks are mandatory.
     5.   The cost is high, relative to other sampling devices.
     6.   Some hydrocarbon solvents may react with the bag.
Sample Direct to Analysis - As the name implies, this technique reduces  the
detainrnent of the sample from the point of extraction to the point  of analysis
to a minimum.  The sample flows from the stack through a sample conditioning
                                     150

-------
system (normally referred to as the interface system)  to the hydrocarbon
detector.  Care must be taken to see that the sample interface  system does
not alter the hydrocarbons to an unknown degree.   Its  operation is  best
checked by periodically introducing a gas of known hydrocarbon  concentration
at the sampling probe.  The previously calibrated analyzer is then  monitored
to verify attainment of the expected response.
     The degree of complexity of an interface system may vary from  the
extreme of simply transporting a portion of the stack gas from the  stack
to the analyzer to the extreme of cooling or heating, filtering, drying,
concentrating or diluting, and reacting the stack gas.  Application of any  -
of these steps are acceptable, provided the nature and extent of their
occurrence is known.
     Some advantages of direct analysis are:
     1.  Time-related degradation phenomena are elimianted.
     2.  The amount of sample equipment is reduced.
     3.  The use of a permanent or semipermanent setup  reduces the frequency
         of leak checks.
     4.  The rapid availability of data may be useful for process control.
     Disadvantages include:
     1,  It is not always possible to provide on-site analytical requirements.
     2.  The method requires a portable or semi portable analyzer, which may
         be an added expense.
     3.  Taking the analyzer to the source means it can only be at  one source
         at a time, whereas a laboratory-situated  analyzer could rapidjy
         analyze samples from several sources.
5. 5    Analysis
     The means of detecting hydrocarbon gases, either directly or indirectly,
that are currently employed are limited to infrared and ionization  spectroscopy
and to mass spectrometry.  Principle variations of these techniques are now
discussed.                           151

-------
    INFRARED
     SOURCE
REFERENCE*
  CELL
              o DO
               £D<>
 o rr
  o  o
  on  o
 0°D°0
  °DO

 °o°o°

 °°0°
 O D  0
  o oo
"-  D   •
    D
    n
                         CHOPPER—*     CD
                                   I—  CD  -

                                    o_no_
                         SAMPLE
                         CELL
                              -^
                                \^
                                     o » o
                                     o _°^
                                     o J
                                    o^-o01
                                      P   o
                         DIAPHRAGM
                         DISTENDED
                                ED-7

                                *  \  '   '   -^
     DETECTOR
                                                 SAMPLE IN
                                                SAMPLE OUT
RECORDER
 SIGNAL
                                        COMPONENT
                                        OF INTEREST
                                          OTHER
                                        MOLECULES
                  CONTROL UNIT
Figure 5-5.  Example of nondispersive infrared absorption analyzer.^
                         152

-------
Nondispersive Infrared Absorption  (NDIR)  -  This  procedure,  the  only  type  of

infrared analysis that has  had much  application  in  this  field,  is  described
                                                         13
in a recent survey of hydrocarbon  methodology  as  follows:

          Nondispersive infrared spectometry is  a technique
     based upon the broadband absorption  characteristics of
     certain gases in the wavelength region of a few micrometers.
     Infrared radiation is  typically directed  through two separate
     absorption cells — a  reference cell  and  a  sample cell (see
     Figure 5-5).  The sealed reference cell is  filled with non-
     absorbing gas, such as nitrogen or argon.  The sample  cell
     is physically identical to the  reference  cell  and receives
     a continuous stream of the gas  being analyzed.  When a
     particular hydrocarbon is present, the IR absorption is
     proportional to the molecular concentration of that .gas.
     The detector consists  of a double chamber separated by
     an impermeable diaphragm.  Radiant energy passing through
     the two absorption cells heats  the two portions of the
     detector chamber differentially.  The pressure difference
     causes the diaphragm to distend and vary a capacitance which
     is measured electronically.  The variation  in  capacitance  is
     proportional to the concentration of the component of  gas
     present.  By optically chopping the IR radiation, the  capaci-
     tance may be made to change periodically, and  as a result,
     the electronic readout problems are facilitated.
          Beckman, Horiba,  and Mine  Safety Appliances all manufacture
     NDIR analyzers based on the above principles,  and use  two  IR
     sources.  Ecological Instruments used a single radiation source.
     Infrared Industries uses two  concave mirrors,  thus allowing a
     single source arrangement.  Bendix produces an analyzer with
     the two detector chambers in  series; both detectors are filled
     with the gas under measurement.  The gas  in the forward chamber
     is heated by the center of the  absorption band; the gas in the
     rear chamber by the edges of  the band.  Hydrocarbon gas in the
     sample will absorb primarily  in the center of  the band and thus
     cause the front chamber to become cooler.  The pressure change
     is detected as a change in capacitance and read out as previously
     described.

          NDIR instruments  are usually subject to interference
     because other gases (e.g., H^O, COp) absorb at the wavelength
     of the gas of interest.  Efforts to eliminate  the interferences
     by use of reference cells or optional filters  are only partially
     successful.  For HC monitoring the detector is filled with one
     or several different hydrocarbons, which  may  be different from
     the HC contained in the sample.  This will  cause a dispropor-
     tionate response.  Other sources of errors include such things
     as gas leaks in detector and  reference cells,  inaccurate  zero
     and span gases, non-linear response, and drift in the  electronics.
                                     153

-------
     Infrared analyzers are also used to measure  carbon  dioxide  produced  by

oxidation of hydrocarbons, which can  then be translated  into  an  indirect

measurement of total  hydrocarbons.

Flame lonization Detection (FID) -  First reported as  a hydrocarbon  detector

by Andreatch and Feinland   and others   approximately 15 years  ago,  the

flame ionization detector (FID) has since achieved considerable  recognition.

FID is also well documented in reference 13:

          With standard FID, air sample is introduced into a
     hydrogen flame (see Figure 5-6).  The combination of even
     0.1 ppm of a hydrocarbon produces measurable ionization
     which is a function of the number of carbon  ions present.
     A collector surrounding the flame is made positive  by an
     external power supply and the  ion current caused by the
     hydrocarbons is  measured electronically.  Since  pure hydrogen
     burning in air produces very little ionization,  the effects
     of background subtraction are  minimized.  The output current
     calibrated in ppm (or percentage) is read on a panel meter
     or chart recorder.
          Hydrocarbons containing nitrogen, oxygen, or halogen
     atoms give a reduced response.  Thus, FID hydrocarbon
     analyzers are almost universally calibrated  in terms of
     a gas such as methane or hexane and the output read in ppm
     of carbon measured as methane^or hexane.
          It is important to note that nitrogen,  CO,  and CO^
     do not produce interferences.   Patterson and Henein point
     out that although there is a very low sensitivity to water
     vapor, condensed water vapor may block the sample entry
     tube and cause erratic readings.  Also when  oxygen  is
     present in excess of 4%, a significantly lower output
     reading may occur.  Beckman Instruments report that
     relative response of the Model 400 Hydrocarbon Analyzer
     to various hydrocarbons, including those with attached
     oxygen, chlorine, and nitrogen atoms in Table 5-1.   The
     response is given in effective carbon numbers (ECN) where:
                            Instrument response caused
                      _„,. _ by atom of given type	
                            Instrument response caused
                            by aliphatic carbon atom

     These values are true for one mode of operation of  a
     specific detector under specific conditions  (e.g.,
     mixed No, H~ fuel).  However,  these numbers  may vary
     widely for different operating conditions and for
     different detectors.

                                     154

-------
o.
                         o
                         O

                         Q)


                         (D
                         (O
                         ub

                         I
                         O)
155

-------
Table 5.1   APPROXIMATE EFFECTIVE CARBON NUMBERS
                                               13
Type of atom
Carbon
Carbon
Carbon
Carbon
Carbon
Carbon
Oxygen
Oxygen
Oxygen
Oxygen
Chlorine
Chlorine
Nitrogen
Occurrence
In aliphatic compound
In aromatic compound
In olefinic compound
In acetylenic compound
In carbonyl radical
In nitrate
In ether
In primary alcohol
In secondary alcohol
In tertiary alcohol, ester
As two or more chlorine
atoms on single aliphatic
carbon atom
In olefinic carbon atom
In amine
Effective
carbon number
+1.0
+1.0
+0.95
+1.30
0.0
+0.3
-1.0
-0.6
-0.75
-0.25
-0.12 each
+0.05
Value simila
                                                     to that for
                                                     oxygen atom
                                                     in corresponding
                                                     alcohol
                       156

-------
Gas Chromatography (GC) - Measurement of hydrocarbons by GC is a well-
                   1 fi-1 R
documented process.  ~    Normally thought of as a device for separating

a hydrocarbon mixture into its individual compounds, GC may also be used

to separate nonmethane hydrocarbons from the other gases in a sample.

     Here again, reference 13 provides a well-stated description of the

basic GC operation:

          Gas chromatographs have been used manually by research
     laboratories to monitor hydrocarbons for many years.  The
     great: power of this technique is the unique ability to separate
     hydrocarbons into a number of individual compounds.-  In prin-
     ciple,GC is a method for physically separating a gaseous mix-
     ture into its components by passing it through a column with a
     high surface-to-volume ratio (see Figure 5-7).  The surface area
     consists of a solid material or a liquid dispersed on a solid.
     The segregation of the various components depends upon their
     selective absorption into the column material.  An inert carrier
     gas moves the sample through the column.  If the gas sample
     consists of different hydrocarbons, the different components
     will require different times to pass through the column.  The
     weakly absorbed components are the first to emerge from the
     column.  The selective process is highly temperature-sensitive
     and thus requires that most of the components of the chromato-
     graph be housed in a temperature-controlled oven.  As the
     various components emerge from the column, their identification
     and concentration are determined by an appropriate detector.
     For hydrocarbons, flame ionization detection, described in the
    .previous section, is almost universally employed.

     A value for total nonmethane hydrocarbons can be generated chromato-

graphically by summing the individual concentrations as they are revealed

by their respective chromatogram peaks.  Another approach is to operate

the chromatograph in such a manner that all nonmetbane hydrocarbons elute

as one from the column.  One peak on the chromatogram then represents total

nonmethane hydrocarbons.

                                     157

-------
                   CARRIER

                     GAS
REGULATOR
tn
OD
                                                                         OVEN
                                                            INJECTOR
                                                                       COLUMN
                                                                                 DETECTOR
                                                                                               TO EXHAUST
                                                                                 RECORDER
                                  Figure 5-7.  Main components of gas chromatograph.

-------
Mass Spectrometry (MS)^ - Mass spectrometry is described as "... essential

to the absolute identification of organic components..." in a recent

                                           19
survey of hydrocarbon detection techniques.     Because individual  compound

identification is not an absolute requirement for the measurement  of total

nonmethane hydrocarbons, MS is mentioned here for the benefit of those

who sample sources emitting single species of hydrocarbons for which

confirmation of their presence is required.

     Mass spectrometers sort a gas into its  individual components  by first

ionizing the individual gas molecules.  The  ionized molecules are  then

subjected to various physical effects that serve to classify or resolve

them into their respective mass numbers.  Several types of mass spectrometers

                       20
are explained by Ewing.    By studying the relative intensities of the mass

numbers generated, a value for total methane hydrocarbons could probably be

generated.


Total Combustion Analysis (TCA) - Personnel  of the Los Angeles County Air
Pollution Control District have reported their progress with implementation

                      for

                      24
                                        21 -24
of the concept of TCA for several years.       The following quotation is
taken from Solo et al.

     The Los Angeles Rule 66 solvent control program has been in use
     for almost a decade.  Measurements of source emissions for
     organic carbon are a vital adjunct for determining compliance
     with the law.  The latter is accomplished using an analytical
     system termed Total Combustion Analysis (TCA).  It involves
     the measurement of stack flow volume, and the collection
     of samples in a stainless steel freezeout trap followed by an
     8-liter stainless tank in tandem.  An analysis of the organic
     carbon in trap and tank is made by conversion and measurement
     of C02.  The results are scaled to the total exhaust flow
                                     159

-------
     volume.  Thus» a measurement of total  organic carbon  in  pounds
     per hour is obtained, which for practical  purposes  is inter-
     preted as being very close to the actual  organic quantity.

     The detector system has historically been  a non-dispersive
     infrared analyzer sensitized to CCk-  Recently,  we  have
     been testing a hydrogen flame ionization  detector (FID)  in
     place of the non-dispersive system.   In the FID  version  of
     the TCA instrument, it is necessary  to  convert  the CCL  to
     CH^ in order to utilize the enhanced sensitivity of the
     detector.  This is accomplished over a nickel catalyst.
     Results indicate that the newer FID  detector is  preferable
     for low concentration cases, and can be used satisfactorily
     in any event for all measurements of organic carbon.

     Figure 5-8 illustrates the analytical  steps for  both  the cold trap

and the evacuated cylinder.  Other sampling devices,  such  as  the Universal

Collector, syringe, purge flask, arid collapsible bag  could also  be used.

     Because all hydrocarbons reach the FID as  CH., the  problem  of varying

response factors, as explained earlier, is eliminated.

     Corrosive gases created by combustion of halogenated  hydrocarbon could

create problems for this system.  A possible remedy would  be  to  install a

selective absorber for these gases between the oxidizer  and the  reducing

catalyst.

     No commercial, fieldproven TCA's are presently available.  If a

semi portable version can be developed, this approach  will  become even more

appealing, because the direct sample technique can then  be used.  Substitutions

of an FID in place of the NDIR greatly enhances the analysis  system's

sensitivity.


5.6  Conclusion

     This section has dealt with the determination of total nonmethane hydro-

carbon emissions through emission measurements.  In the  course of describing

various sampling and analytical techniques that are currently available,


                                    160

-------
TRAP ANALYSIS MODULE
      SAMPLE
       TRAP
   (CONDENSABLES)
   TRAP FURNACE
       (600°C)
     PLATINUM
      OXIDIZER
       (850°C)
     C02   H20
    WATER TRAP
       (-78 °C)
     INFRARED
     DETECTOR
   INTERMEDIATE
    COLLECTION
      VESSEL
    ALIQUOT C02
    FOR ANALYSIS
        TANK ANALYSIS MODULE
            SAMPLE TANK
               (GASES)
                                                   T
                                             SAMPLE ALIQUOT
                                                                 HELIUM
                                                              -CARRIER GAS
                                            CHROMATOGRAPHIC
                                                 COLUMN
CONTROL VALVE
                                                   CO
                                                  'CH4
                                                  C02
CONTROL VALVE
B
A
C
K
B
L
U
S
H
              HEPCALITE
               OXIDIZER
                (8SO°C)
                                               C02 1 H20
             WATER TRAP
                (-78 °C)
                                              co
                                                             CONTROL VALVE
                                                               HYDROGEN
                                              NICKEL CATALYST
                                                  (400°C)
       PREVIOUS NDIR
          SYSTEM
                                      AIR
                                               NEW FID SYSTEM
   Figure 5-8. Block flow diagram of TCA trap and tank analysis systems
   showing difference between NDIR and ID detection.24
                            161

-------
concepts have been presented that should orient those persons  responsible
for determining source compliance with emission standards  to the type of
questions that must be resolved before a particular method is  applied.
                                     162

-------
5.1 REFERENCES FOR CHAPTER 5

     1.   Air pollution.  2nd Ed.,  Vol.  II.  Stern,  A.C.  (ed.). New York,
         Academic Press,  1968.  p.3-54.

     2.   Sampling Atmospheres  for Analysis of  Gases  and Vapors. American
         Society for Testing  and  Materials, Philadelphia,  Pa., DI605
         ASTM Annual Standards. Part  26,  1974.  p.  285-306.

     3.   Sampling and Storage of  Gases  and Vapors, Methods of Air Sampling
         and Analysis. Part 1.  Intersociety Committee  of American Public
         Health Association,  Washington,  D.C. 1972 p. 43-55

     4.   Title 40 - Protection of Environment.  Test  Methods  1-4. Federal
         Register. 36_ (247):  24882-24887,  December 23, 1971.

     5.   Air Sampling Instruments. 4th  Edition American Conference  of
         Governmental and Industrial  Hygienists,  Sections  A  and B.  1972.

     6.   White, L.D., D.G. Taylor, P.A. Mauer, and R.E. Kupel.  A Convenient
         Optimized Method for the Analysis of  Selected Solvent Vapors  in
         the Industrial  Atmosphere. Amer.  Ind.  Hyg.  Assoc. J. 31(2): 225,
         March - April 1970.

     7.   Mueller, F.X.,  and J.A.  Miller.  Determination of  Organic Vapors
         in Industrial Atmospheres. Amer.  Lab.  6(5): 49-61,  May 1974.

     8.   Isbell, A.F., Jr. Development  of Selective  Hydrocarbon Sampling
         System and Field Evaluation  with Conventional Analytical System.
         U.S. Environmental Protection  Agency, Research Triangle Park,
         N.C. under Contract  No.  68-02-1201. Publication No. 650/2-75-050.
         August 1975.

     9.   Gadomski, R.R., A.V.  Gimbrone, M.P. David,  and W.J. Green.
         Evaluations of Emissions and Control  Technologies in the Graphic
         Arts Industries. Phase II: Web Offset and Metal Decorating Processes.
         Graphic Arts Technical  Foundation, Prepared for U.S. Environmental
         Protection Agency, Research  Triangle  Park,  N.C. under Contract No.
         68-02-001. Publication No. APTD-1463.  May 1973.

    10.   Hales, J.M., and N.S. Laulainen.  Report  on  Verification of Halogenated
         Hydrocarbon Testing and  Monitoring Methodology. Battelle-Northwest,
         Rich!and, Washington. Prepared for U.S.  Environmental Protection
         Agency, Research Triangle Park,  N.C.  under  Contract No. 68-02-1409
         (Task 15). 1975.

    11.   Title 40 - Protection of Environment. EPA Method  106  - Determination
         of Vinyl Chloride from Stationary Sources.  Federal  Register-40(248):
         59550, December 24,  1975.
                                     163

-------
12.  Schuetzle, D.,  T.  J.  Prater, and S.  R.  Ruddell.  Sampling  and
     Analysis of Emissions from Stationary  Sources  1.   Odor  and Total
     Hydrocarbons.  J.  Air  Pollution  Control  Assoc.  25(9):  925-32,
     September 1975.

13.  Air - Hydrocarbon Monitoring Instrumentation.   Lawrence Berkeley
     Laboratory, University of California,  Berkeley,  California.
     November 1973.

14.  Andreatch, A.J.  and R. Feinland. Continuous  Trace  Hydrocarbon
     Analysis by Flame lonization.  Anal.  Chem.  32(8): 1021-1024,
     July 1960.

15.  Morris, R.A.,  and R.  L. Chapman.  Flame lonization Hydrocarbon
     Analyzer. J. Air Pollution Control  Assoc.  11(10):  467-469. October
     1961.

16.  McNair, H.M.,  and E.  J. Bonelli. Basic Gas Chromatography.
     Varian Aerograph, Inc., Walnut  Hills,  California.  1969.

17.  Total Hydrocarbons by GC-FID.   In:  Methods Of  Air  Sampling and
     Analysis.  Intersociety Committee of American  Public  Health
     Association, Washington, D.C.  ISC 41301-02 71T.  1972

18.  Jeltes, R. and E. Burghardt. Automatic Gas Chromatographic Measure-
     ment of C-, - Cj-  Hydrocarbons in Air. Atmos.  Environ.  6: 793-805,
     1972.     '    b

19.  Survey of Methods for Measuring Total  Hydrocarbons. DeBell and
     Richardson, Inc., Enfield, Conn. Prepared  for  U.S. Environmental
     Protection Agency, Research Triangle Park, N.C., under  Contract
     No. 68-02-2075.  1976.

20.  Ewing, G.W. Instrumental Methods of Chemical Analysis.  New York,
     McGraw-Hill Book Company, 1969.

21.  MacPhee, R. D.,  and M. Karamoto.  Recommended  Test Methods for
     Organic Solvents and  Vapors (Rule 66). Air Pollution  Control
     District, County of Los Angeles, Los Angeles,  California.
     April 1968.

22.  DeVorkin.H.Sampling for Compliance with Rule 66. (Presented  at
     62nd Annual Meeting of Air Pollution Control Association. New York.
     Paper No. 69-49. June 22-26, 1969.)

23.  Salo, A.E., \L  L. Oaks, and D.  MacPhee.   Measuring the Organic
     Carbon Content of Source Emissions for Air Pollution  Control.
     (Presented at  67th Annual Meeting of Air Pollution Control Assoc.
     Denver.  Paper No. 74-190. June 9-13, 1974.)
                                 164

-------
24.   Sale, A.E., S.  Witz, and Robert D.  MacPhee.  Determination of
     Solvent Vapor Concentrations by Total  Combustion Analysis:   A
     Comparison of Infrared with Flame lonization Detectors.  (Presented
     at 68th Annual  Meeting of the Air Pollution  Control  Association.
     Boston. Paper No.  75-33.2 June 15-20,  1975.)

25.   Neal, R.C., P.  L.  Hayden. D.R. Grove and E.A.  Brackbill.   Test
     Methods for the Evaluation of Organic  Emissions. (Presented at
     1st Annual Conference of Source Evaluation Society.  Dayton.
     September 17-19, 1975.)

26.   1-Joods, J. and R. Melcher. Review of Analytical  Methods  and
     Techniques - (Study to Support New Source Performance Standards
     for Industrial  Degreasing Operations).   Dow  Chemical  Co., Freeport,
     Texas - Prepared for U.S. Environmental  Protection Agency, Research
     Triangle Park,  N.C.,under  Contract No.  68-02-1329.  (Sub  Task 2)
     1975.
                                165

-------
I
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 i. REPORT NO.
  EPA-450/2-76-028
                             2.
                                                           3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Control  of Volatile Organic Emissions  from Existing
Stationary Sources - Volume 1, Control  Methods  for
Surface  Coating Operation
             5. REPORT DATE
               November,  1976
             6. PERFORMING ORGANIZATION CODb
7. AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO.
                                                           OAQPS  No.  1.2-067
9. PERFORMING ORGANIZATION NAME AND ADDRESS
J.S.  Environmental Protection Agency
Office  of Air and Waste Management
Dffice  of Air Quality Planning and Standards
Research  Triangle Park, North Carolina  27711
                                                           10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.
 12. SPONSORING AGENCY NAME AND ADDRESS
                                                           13. TYPE OF REPORT AND PERIOD COVERED
                                                           14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
     Available methods which can be  used  to control the emission  of  organic vapors
 from surface coating operations are  described.   The methods consist  of two types:
 (1)  add-on  control equipment, and  (2)  process and material changes.   Available add-
 on equipment includes direct-flame  incinerators, catalytic incinerators,  and activated
 carbon  adsorbers.   Process and material changes which reduce or eliminate the use of
 organic solvents include (a) water-borne  coatings, (b) high solids coatings, (c) powder
 coatings,  (d)  hot  melt formulations,  (e)  electrostatic spraying,  (f)  electron beam
 curing, (g)  ultraviolet curing.  Graphs are given to determine the cost of incinerators
 at varying volumes and variation in  inlet temperature, vapor concentration, degree of
 heat recovery, fuel costs, and hours  of operation.  Graphs are given  to determine
 the cost of  carbon adsorbers under  varying volumes and vapor concentration.  The
 available methods  'of measuring volatile organic emissions are discussed.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                          c.  COSATI Field/Group
 Air Pollution
 Surface  Coating Operation
 Vapor  incinerators
 Carbon adsorbers
  Air Pollution Control
  Stationary Sources
  Organic vapors
IS. DISTRIBUTION STATEMENT
                                              19. SECURITY CLASS (This Report)
                                                Unclassified
                                                                         21. NO. OF PAGES
                              166
 Unlimited
20. SECURITY CLASS (Thispage)
  Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
                                             166

-------
ENVIRONMENTAL PROTECTION AGENCY
     Technical Publications Branch
       Office of Administration
Research Triangle Park, North Carolina 27711
        OFFICIAL BUSINESS

   AN EQUAL OPPORTUNITY EMPLOYER
     POSTAGE AND FEES PAID
ENVIRONMENTAL PROTECTION AGENCY
           EPA • 335
                                                                           SPECIAL FOURTH-CLASS RATE
                                                                                   BOOK
                              Return this sheet if you do NOT wish to receive this material C_J
                              or if change of address is needed I...  1. (Indicate change, including
                              ZIP code.)
                            PUBLICATION NO. EPA-450/2-76-028
                                        (OAQPS NO.  1.2-067)

-------