CORPORATION
FINAL
EPA CONTRACT
TASK
REPORT
NO. 68-02-1319
NO. 46
EVALUATION OF A CARBON ADSORPTION/
INCINERATION CONTROL SYSTEM
FOR AUTO ASSEMBLY PLANTS
-------
RADIAN
CORPORATION
Radian Project 200-045-46
FINAL REPORT
EPA CONTRACT NO. 68-02-1319
TASK NO. 46
EVALUATION OF A CARBON ADSORPTION/
INCINERATION CONTROL SYSTEM
FOR AUTO ASSEMBLY PLANTS
Submitted to:
Environmental Protection Agency
100 California Street
San Francisco, California 94111
Attention: Mr. Fred Thoits, Project Officer
25 May 1976
Prepared by:
E. C. Cavanaugh
G. M. Clancy
R. G. Wetherold
8500 Shoal Creek Blvd /P.O. Box 9948/Austin, Texas 78766/(512)454-4797
-------
RADIAN
CORPORATION
ABSTRACT
The results of work performed by Radian Corporation under
EPA Contract No. 68-02-1319 (Task 46) are presented in this report.
e purpose of the work was to determine the technical feasibility
and the cost of using carbon adsorption/incineration systems to
reduce auto assembly plant hydrocarbon emissions to comply with
EPA solvent regulations^
The carbon adsorption/incineration system appears, in
general, to be technically feasible and applicable for the control
of hydrocarbon emissions from auto assembly plant paint spray
booths and paint baking ovens. The solvent removal costs are in
the ranges of $1020 - 1600/ton for paint spray booth emissions
and $540 - 1380/ton for paint baking ovens. These costs are
computed for effluent air streams containing averages of 50 ppm
and 300 ppm by volume of toluene for paint spray booths and bake
ovens, respectively. The costs were based on the equipment and opera-
ting expenses for a battery limits plant only. The cost of off-site
facilities, which can be substantial, are not included in the solvent
removal costs.
Installation costs for battery limit plants range from
$0.5 MM to $4.7 MM for units to treat from 60,000 to 900,000 scfm
of solvent-laden air from paint booths, and from $0.4 MM to $3.3
MM for units to treat 10,000 to 250,000 scfm of air from paint
bake ovens.
A typical auto assembly plant was defined and a system
was designed to treat the solvent-laden air to comply with solvent
emission regulations such as Los Angeles' Rule 66 (3000 pound per
day from each spray booth and an 857q reduction in solvent emissions
from each oven). The estimated capital cost of the battery limits
control system was $4.5 MM, and the solvent removal cost was
-------
RADIAN
CORPORATION
determined to be $0.71 per pound of solvent removed. The yearly
operating cost was estimated to be $1.47 MM for the battery limits
plant.
Upon completion of the final draft of this report, a
workshop was held on February 26, 1976 at the EPA Region IX offices
in San Francisco. The purpose was to critique the report and to
discuss the applicability of carbon adsorption systems for the
treatment of auto assembly plant air streams. Representatives from
various industries and governmental agencies attended. At the
conclusion of the workshop written comments were invited from all
the attendees. Those comments which were received have been in-
cluded in Appendix A of this report.
-------
RADIAN
CORPORATION
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
2.0 CONCLUSIONS 3
3.0 ORGANIC SOLVENTS: EMISSIONS AND CONTROLS 9
3.1 Types of Coatings 12
3.2 Solvents 12
3.3 Organic Solvent Emission Regulations 14
3.4 Auto Assembly Plant Paint Opertions 18
4.0 CARBON ADSORPTION SYSTEMS: PROCESS DESIGN 22
4.1 Design Considerations 22
4.1.1 Types of Adsorbers 23
4.1.2 Process Variables 26
4.1.3 Carbon Regeneration 33
4.2 Paint Spray Booth Solvent Emission Control. 38
4.2.1 Description of System 38
4.2.2 Equipment Description 39
4.2.3 Case Studies 46
4.3 Paint Oven Solvent Emission Control 47
4.3.1 Description of System 47
4.3.2 Equipment Description 50
4.3.3 Other Case Studies 53
4.4 Emission Control System for a Typical
Auto Assembly Plant 53
4.4.1 System Description 54
4.4.2 Description of Equipment 56
4.4.3 Plot Plan 61
ii
-------
RADIAN
CORPORATION
TABLE OF CONTENTS (Cont.)
Page
5.0 PROCESS ECONOMICS 63
5.1 Cost Data 63
5.1.1 Equipment Costs 63
5.1.2 Utility Costs 65
5.2 Paint Spray Booth Emission Control Systems. 66
5.3 Paint Oven Emission Control Systems 70
5.4 Control System for a Typical Auto Assembly
Plant 75
5.5 Sensitivity Study 82
6.0 ENERGY CONSIDERATIONS 90
6.1 Fuel Requirements 90
6.2 Fuel Availability 91
6.3 Energy Recovery 91
7.0 OTHER CONSIDERATIONS 94
7.1 Plant Size/Land Requirements 94
7.2 Plant Operations 95
8.0 SUMMARY 96
REFERENCES 97
APPENDIX A Comments from Workshop Participants . 99
iii
-------
RADIAN
CORPORATION
1.0 INTRODUCTION
The emission of organic solvent vapors to the atmosphere
from operations and processes in which these solvents are used
(rather than manufactured) is significant. In studies in the
Los Angeles and San Francisco Bay areas, the contribution of
organic solvent vapors represents as much as 15-207o of the total
organic emissions (CH-275, WO-070). In a 1963 study in the San
Francisco Bay Area (WO-070) solvent emissions from surface coating
operations made up over half of the total organic solvents issuing
to the atmosphere.
Automobile assembly plants are the source of a large
volume of organic solvent emissions from paint spray booths and
paint baking ovens. An auto assembly plant may emit from 20,000
to 40,000 pounds of organic solvents per day. Large quantities
of air are passed through paint spray booths in order to minimize
worker exposure to paints and solvents. Thus, while significant
quantities of solvents are emitted, they are present in small
concentrations in large amounts of air. Typically, concentra-
tions of organic solvents are in the range of 30-300 vppm (ppm by
volume) in air from paint booths and 100-500 vppm in the air emitted
from paint baking ovens.
In recent years, laws have been enacted requiring reduc-
tions in the amount of organic solvents emitted from various sources
to the atmosphere. The application of these laws to automobile
assembly plant paint operations presents some unique problems
because of the very large quantities of air, containing low con-
centrations of solvents, which must be treated to remove the
organic materials. Incineration of some assembly plant air streams
has been used, but fuel cost and availability preclude any future
large-scale use for high volume streams containing low concentrations
of solvents.
-------
RADIAN
CORPORATION
The use of carbon adsorption systems, with incineration
of the desorbed solvents, has been proposed. The purpose of this
study was to determine the cost of installing, operating, and
maintaining a carbon adsorption/incineration control system for
a typical auto assembly plant using lacquer and enamel paints and
employing baking ovens. The study includes an investigation of
the fuel cost for adsorption/incineration, energy recovery and
utilization alternatives, physical size requirements, and the
sensitivity of the cost of the system to various important param-
eters such as solvent concentration, unit size, and air velocity
through the carbon beds.
Conceptual process designs were developed for carbon
adsorption/incineration systems to treat air from paint spray
booths and paint baking ovens. The costs of these units were
determined as a function of the quantity of air to be treated.
A typical plant was defined, and the types and sizes of carbon
adsorption/incineration systems for this plant were selected.
The costs of installing, maintaining, and operating the plant were
determined, and the physical size of the plant was estimated.
A workshop was held on February 26, 1976 at the EPA Region
IX offices in San Francisco, California. Automobile manufacturers,
carbon adsorption equipment manufacturers, and various local, state,
and federal environmental agencies were represented. The purpose
of the workshop was to discuss the suitability of carbon adsorption
systems for use in auto assembly plants and to critique this report.
At the conclusion of the workshop, all the participants were invited
to submit written comments which would be included in the final
version of this report. Several of the attendees did respond, and
their comments are included in Appendix A of this document.
The observations are quite detailed and extensive. They
represent the viewpoints and reflect the experience of both the
automobile companies and the carbon adsorption equipment manufacturers.
Their comments are a valuable addition to this report.
-2-
-------
RADIAN
CORPORATION
2.0 CONCLUSIONS
The results of this investigation support the following
conclusions:
1) Technical Feasibility
In general, the use of carbon adsorption/incin-
eration systems for the effective removal of
organic solvent vapors from air streams being
emitted from paint spray booths and paint baking
ovens appears to be technically feasible. The
term "technically feasible" is used to indicate
that no major technological breakthroughs are
required to accomplish the removal of solvent
vapor from air as described in this report. There
may be some instances in which certain solvents
might be difficult to remove by carbon adsorp-
tion or in which a solvent, once adsorbed,
might be difficult to remove from the carbon
using steam regeneration.
2) Cost of Paint Spray Booth Emission Control
The cost of purchasing and installing a
carbon adsorption/incineration system for
removal of solvents from spray paint booth
air ranges from $0.52 MM for a unit to treat
60,000 scfm of air to $4.7 MM for a plant
processing 900,000 scfm. These costs are
for battery limits plants installed at
ground level in mid-1975. The air stream
is assumed to contain 50 ppm (by volume) of
toluene. The accuracy of these costs is
estimated to be ±30%.
-3-
-------
RADIAN
CORPORATION
The annual operating costs for paint spray
booth solvent emission control vary from
SO.17 MM/year to $1.66 MM/year for the
units processing 60,000 scfm and 900,000
scfm of air, respectively. These costs in-
clude amortization of the capital cost over
a 10 year period at 12% interest. They are
based on equipment and utilities require-
ments for battery limits plants only. The
operating costs for off-site facilities are
not included.
The cost of solvent removal with the above
systems ranges from $0.51/lb for the
900,000 scfm unit to $0.80/lb for the
60,000 scfm plant.
3) Cost of Paint Baking Oven Emission Control
Carbon adsorption/incineration systems for
controlling solvent emissions from paint baking
ovens can be purchased and installed for
$0.4 MM to $3.3 MM for units sized to process
10,000 to 250,000 scfm of air, respectively.
These costs are for battery limits plants
installed in mid-1975 at ground level. The
air from the ovens is assumed to be at 370°F
and to contain 300 ppm (by volume) of toluene.
Again, the accuracy of these costs is esti-
mated at ±307c
The operating costs of these plants range
from $0.12 MM/year for processing 10,000 scfm
of air to $1.48 MM/year for a 250,000
scfm air treatment plant. These costs
-4-
-------
RADIAN
CORPORATION
include amortization of the capital
costs over a 10 year period at 1270
interest. They are based on equip-
ment and utilities requirements for
battery limits plants only. The
operating costs for off-site facilities
are not included.
The cost of solvent removal from the
air emitted from paint baking ovens with
the above systems varies from $.69/lb for
treating 10,000 scfm to $0.27/lb for
processing 250,000 scfm of air containing
300 ppm of toluene.
4) Cost of Treating Air From a Typical Plant
In a typical auto assembly plant contain-
ing four paint booths and 8 baking ovens which
are emitting excessive amounts of solvent
vapors, the cost of purchasing and installing
a carbon adsorption system is about $4.5 MM.
The system is sized to treat 570,000 scfm of
solvent-laden air. This cost is for a battery
limits plant installed at ground level in
mid-1975.
The cost of operating this system is about
$1.5 MM per year. For a plant operating with two
shifts and producing 80 units per hour, the
operating cost is equivalent to about $3.50 -
$4.00 per car. This is equivalent to a solvent
removal cost of $0.71 per pound.
-5-
-------
RADIAN
CORPORATION
This recovery system is designed to treat only
an amount of effluent air sufficient to reduce
solvent vapor emissions below a level of 3000
lb/day of solvent from each spray booth. The
solvent emissions from ovens are reduced by 85%.
The air to be treated amounts to about 40% of
the total air emitted to the atmosphere.
In some cases, it may be possible to reduce
the amount of air to be treated by processing
the air streams containing high solvent con-
centrations and rejecting those of low concen-
trations .
5) Size of System
Carbon adsorption systems of sufficient size to
treat more than about 30% of the air emitted
from typical auto assembly plants are larger
than any regenerable carbon adsorption system
now in existence. The design, construction, and
operation of a system of this size requires some
extrapolation of existing technology. If widely
applied, some strain on material and equipment
supplies might be experienced.
6) Solvent Recovery
The recovery of solvents from the air (as con-
trasted to incineration) is not practical because
of the number of solvents present and the diffi-
culty involved in separating them into components
of sufficient purity to be acceptable for reuse.
-6-
-------
RADIAN
CORPORATION
7) Steam Regeneration
Steam was selected as the desorbing or regener-
ating medium. If the amount of steam required
to regenerate the carbon is equivalent to or less
than 15 pounds of steam per pound of adsorbed
solvent, the amount of heat generated from incin-
erating the solvent is sufficient to produce the
necessary steam. Some additional fuel is required,
however, for heatup and control purposes.
8) Heat Recovery
The amount of heat produced in the incineration
of solvent is normally more than that required
for the production of regeneration steam. This
excess heat can be recovered either as additional
steam or by direct exchange to heat air for the
paint baking ovens. Credit for the recoverable heat
was included when determining the operating and
solvent removal costs for the cases presented in
this report.
9) Land Requirements
The physical size of the required carbon adsorp-
tion control systems is quite large. A central
system large enough for a typical auto assembly
plant would require about 0.5 - 1.0 acres of
land.
10) Retrofit of Existing Plants
In retrofitting existing assembly plants with
carbon adsorption systems, the location of
-7-
-------
RADBAN
CORPORATION
existing paint booth and oven exhaust ducts with
respect to a location available for the adsorption
systems will be very important. The cost of col-
lection ductwork and, if necessary, additional
blowers could be substantial. In some cases the
cost of off-site facilities might approach the cost
of the battery limits plant. Since this equipment
is outside the battery limits of the carbon adsorp-
tion system, the costs are not included in the
operating and solvent removal costs.
11) Pilot Studies
In order to determine the applicability of carbon
adsorption systems for removal of solvents in
specific air streams and to obtain design informa-
tion, pilot studies will be necessary. These could
be accomplished using small package adsorption units.
-8-
-------
RADIAN
CORPORATION
3.0 ORGANIC SOLVENTS: EMISSIONS AND CONTROLS
The emission of organic solvent vapors from stationary
sources contributes significantly to the total organic emissions
from all sources. For example, in the Los Angeles and San
Francisco Bay areas emissions from solvent usage industries rep-
resent from 13 to 22 percent of the total organic emissions
(CH-275, WO-070). The amounts of emissions are shown in Table
3.0-1.
It is generally agreed that surface coating is by far
the largest end use of solvents, ranging from 30 to 60 percent
of the total usage. Solvent consumption by the surface coating
industry has been estimated to be from 6 to 8 billion pounds per
year (MS-001). Of the end use categories, surface coating opera-
tions use the largest quantities of mixed petroleum solvents,
both aromatic and aliphatic, as well as significant amounts of
the oxygenated solvents (alcohols, ketones, esters).
Recent data from the Bay Area Air Pollution Control
District (1969) show that 58 percent of estimated solvent evap-
oration emissions in the District derive from surface coating
operations (MS-001). The bulk of the emissions from surface
coating operations results from spray painting and evaporative
drying or curing operations. This is evident from Table 3.0-2
which shows that about 62 percent of the emissions from the surface
coating industry in the San Francisco Bay Area are derived from
spray booths, ovens, and dryers (MS-004).
-9-
-------
RAD AN
CORPORATION
Table 3.0-1
ESTIMATED EMISSIONS OF ORGANIC VAPORS AND GASES
FROM VARIOUS SOURCES
Emissions in Tons per Calendar Day
Los Angeles San Francisco
County, Bay Area,
Jan. 1962 1963
Solvent Usage 445 280
Motor Vehicles 1300 1100
Petroleum Refining 85 68
Petroleum Product Marketing 100 83
Petroleum Production 60
Others 60 569
Total 2050 2100
SOURCES: CH-275, WO-070
-10-
-------
RADIAN
CORPORATION
Table 3.0-2
AMOUNTS OF ORGANIC POLLUTANTS EMITTED BY SOURCES WITHIN THE SURFACE
COATING INDUSTRY IN THE SAN FRANCISCO BAY AREA (1969)
Organic
Emissions,
Source Tons/Day
Spray booths 87.7
Ovens and dryers 34.1
Flowcoaters, dip tanks,
washers 3.9
Brushing, rolling,
non-industrial spraying 72.0
197.7
Source: MS-004
-11-
-------
RADBAN
CORPORATION
3.1 Types of Coatings
There are five general types of surface coatings. These
are paint, enamel, varnish, lacquer, and shellac. Paints, enamels,
and lacquers are all used in the automobile industry.
Paint is a highly pigmented drying oil thinned with
a low solvency power solvent, referred to as a thinner. The
drying oil is an unsaturated polymer which oxidizes and polymerizes
further in air to form a resinous protective film. It is only
partially soluble in the thinner. Natural drying oils have been
predominantly used, but the shift is now toward the synthetic
types. Enamels are basically the same as paints except that
synthetic drying oils are present in higher concentrations.
Applied enamel and paint coatings dry and cure by
evaporation of the thinner and by oxidation-polymerization of
the drying oil to form the resinous film. In contrast, lacquer
vehicles consist of a resin dissolved in a high solvency power
solvent. After application, the drying occurs by evaporation of
the solvent and deposition of the resin-pigment film. The early
lacquers consisted of a cellulose nitrate resin, while the
present ones are made with acrylic resins.
3.2 Solvents
Compositions of surface coatings and particularly the
amounts of solvents and thinners used vary with the end use and
type of application method. Typical compositions of paints,
enamels, and lacquers are presented in Table 3.2-1. Volatile
organic constituents range from less than 10 percent for high
solids content coatings for dip and roller applications to over
70 percent for spray applications of lacquers.
-------
TABLE 3.2-1
TYPICAL COMPOSITIONS OF PURCHASED SURFACE COATINGS (7»)
Type Nonvolatile Hydrocarbons Esters&
Coating Portion Aliphatic Aromatic Alcohols Ketones Ethers
Paint 44 56 - - -
Enamel 58 10 30 2 -
Lacquer 23 7 30 9 22 9
OJ
I
SOURCE: MS-004
-------
RADIAN
CORPORATION
Water base paints have volatile organic contents on
the order of 2 to 5 percent.
Table 3.2-2 contains a listing of some of the individual
solvents and thinners which have been used in the surface coating
industry. The thinners are aliphatic hydrocarbons, turpentine,
mineral spirits, and naphtha. They are used mainly in paints,
enamels, and varnishes. The solvents are the aromatics, alcohols,
ketones, ethers, and esters. They are used in varying degrees
in enamels and varnishes, but mainly in lacquer and shellac.
3.3 Organic Solvent Emission Regulations
A number of federal, state, and local regulations are
currently in existence to control the amount of reactive and
unreactive organic pollutants which can be emitted to the atmo-
sphere from various sources. Of particular interest in the
control of organic solvent vapor emissions are Rule 66 of Los
Angeles County and Regulation 3 of the Bay Area Air Pollution
Control District. These two similar regulations are concerned
with control of solvent emissions to reduce smog, which is a
particularly serious problem in these control districts.
Of particular interest to this study are the specific
prohibitive rules dealing with organic solvent emissions. These
are illustrated by Rules 66a, 66b, and 66c of the Los Angeles
County regulations which are summarized below.
Rule 66a
According to this rule the amount of organic material
emitted from processes where solvent-containing materials are heat-
cured, baked, or heat polymerized, or where solvents come into
contact with flame, is limited to 15 pounds per day. The limit
-14-
-------
RADIAN
CORPORATION
TABLE 3.2-2
SOLVENTS USED IN THE SURFACE COATING INDUSTRIES
Molar Volume Boiling
(Vm) Point
Compound cm3/mol °F
Dodecane 274 421
Undecane 251 383
Decane 229 345
Butyl Carbitol 213 448
Nonane 207 302
2,6-Dimethyl-4-heptanone 207 334
Diethylcyclohexane 207
Butylcyclohexane 207 354
1-Methylpentyl dcetate 194
Diethylcyclopentane 192 307
Butylcyclopentane 188 314
Octane 185 257
Butylbenzene 185 361
2-Pinene (turpentine) 184 300
Butyl-2-hydroxy-propanate 183
Pentyl acetate 174 300
Carbitol 169 393
Heptane 163 208
Ethylcyclohexane 162 269
Isopropylbenzene 162 306
Propylbenzene 162 318
Dimethylcyclohexane 162 246
2-Butoxy-ethanol 160 340
2-ethoxy-ethyl acetate 160 313
1-Methylpropyl acetate 155
Butyl acetate 152 260
2-Methyl-propyl acetate 155
-15-
-------
RADIAN
CORPORATION
SOLVENTS USED IN THE SURFACE COATING INDUSTRIES (Cont.)
Compound
Diisopropyl ether
4-Methyl-2-pentanol
Dimethyl cyclopentane
Ethyl cyclopentane
2-Hexanone
4-Methyl-2-pentanone
Xylene
Ethylbenzene
Hexane
4-Methyl-3-penten-2-one
Propyl acetate
Isopropyl acetate
Cyclohexanol
1-Pentanol
2-Pentanone
Cyclohexanone
Toluene
Cyclohexane
Diethyl ether
Ethyl acetate
2-Methyl-l-propanol
1-Butanol
2-Butanol
Cyclopentane
1-Nitropropane
2-Nitropropane
Butanone
Benzene
2-Methoxy-ethanol
Molar Volume
(Vm)
cm3/mol
152
150
144
144
141
141
140
140
140
133
129
129
127
127
118
118
118
118
106
106
105
105
105
100
97
97
96
95
92
Boiling
Point
156
266
217
242
291
277
154
266
214
199
321
279
215
231
231
178
94
171
244
210
120
266
248
175
176
253
-16-
-------
RADIAN
CORPORATION
SOLVENTS USED IN THE SURFACE COATINGS INDUSTRIES (Cont.)
Molar Volume Boiling
(Vm) Point
Compound cm3/mol °F
2-Propanol 84 180
1-Propanol 84 206
Methyl acetate 83 135
Nitroethane 75 239
Propanone 74 ^33
Ethanol 61 173
Nitromethane 53 214
Methanol 42 149
2-ethyl hexanol 365
Acetone 133
Methylethylketone 175
Amyl acetate 298
Cellosolve
Cellosolve acetate 314
Ethanol 173
Naphtha (petroleum)
Mineral spirits
Trichloroethylene 189
-17-
-------
CORPORATION
applies regardless of whether the solvents used in the original
material are defined as photochemically reactive or photochemically
nonreactive. If the 15 pounds per day limit cannot be achieved,
the emissions must be reduced by at least 85 percent.
Rule 66b
A limit of 40 pounds per day is placed on the quantity
of photochemically reactive organic material which may be discharged
into the atmosphere in any one day from any one article or machine
which employs, applies, evaporates, or dries such a solvent.
Again, if this limit cannot be reached, an 85 percent reduction
in emissions must be achieved.
Rule 66c
According to this rule the discharge of photochemically
nonreactive organic solvents from any source must be either
reduced by at least 85 percent or reduced to not more than 3000
pounds per day. As will be discussed in later sections of this
report, solvent emission regulations of this type can seriously
impact the painting operations in automobile assembly plants.
3.4 Auto Assembly Plant Paint Operations
The basic painting operations occurring on the assembly
line in automobile and commercial vehicle assembly plants are
generally similar. There are many minor variations which are
dependent on types of vehicles, paints, and assembly sequences.
In a typical operation, the automobile body passes
continuously from the body shop (where the main metal portions of
the body are assembled) to the paint operation. In the paint
-18-
-------
RADIAN
CORPORATION
operation the bodies are coated first with a primer arid then with
a topcoat which may be an enamel or a lacquer. After the body has
been painted it continues on to the trimshop where seats, dash-
boards, linings, etc. are added.
Some or all of the body parts may arrive at the plant
covered with a light coat of oil to prevent rust. After assembly
the body is wiped with solvent to remove the oil and treated in
preparation for painting.
The prime coat may be applied by dip-coating, spray
painting, or a combination of both. An electrophoretic process
(ELPO) may be used in a dip-coating operation where the car body
is passed through a bath containing a water-based primer. As an
alternative, or in addition to the ELPO process, the prime coat
(solvent or water based) may be sprayed onto the car body either
manually or by a combination of manual and automatic operations.
After the primer coat has been applied it is baked onto
the body in the first baking oven, where the car body passes
through a zone of gradually increasing temperature. Heat is
supplied by air which is generally heated by indirect exchange
with hot flue gases from natural gas combustion.
The primed auto body is cooled after leaving the prime
oven and wet sanded to insure adhesion of the topcoat. The body
then continues into the first color booth where the topcoat is
sprayed on. This is done manually or by a combination of auto-
matic and manual spraying. The automobile body is then sent to
another baking oven to bake or cure the topcoat.
-19-
-------
RADIAN
CORPORATION
There are other possible sequences and combinations of
paint booths and ovens, but the above procedure is generally
common to most operations. Other paint spray booths along the
assembly line might include tutone booths, parts booths, or repair
booths. An assembly plant may contain two or more assembly lines,
each with its own series of paint spray booths and baking ovens.
Spray booths are operated with air rates of at least
100 ft3/min per square foot of booth opening as required by various
state and federal laws (MS-004). The flow of air is normally in a
vertical downward direction. In some cases (small parts, etc.)
where paint can be applied from a single spray source, air may be
directed in a horizontal direction away from the spray source.
Because of overspray of the paint, particulates consisting of pig-
ments, drying oils, and resins are present in the air exhausted
from the paint booths. This air is normally passed through a water
wash system to remove the particulates. These particulates are
subsequently removed from the water by filtration. In general, the
water wash system does not remove significant quantities of solvents
from the air.
With the required air rates and the large size of
many spray booths, it is evident that very large quantities of
air pass through the booths. This air contains solvents in con-
centrations which are generally in the range of 30-300 vppm
(RO-227, CA-288). The volume of air passing through a large spray
booth may be as high as 300,000 - 450,000 scfm. Spray booths may
have as many as 10 or more exhaust ducts. Thus, even though
solvent concentrations may be low, the total amount of solvent
emitted may easily exceed the maximum allowable level of 3000
pounds per day as specified in many regulations.
Air volumetric flow rates through the baking ovens are
lower than those found in spray booths. The amount of air supplied
to the oven is governed by requirements for temperature control
-20-
-------
RADIAN
CORPORATION
and solvent concentrations in the oven. The temperature must
permit proper curing and the solvent concentration must remain
below the lower explosive limits. Depending on the type of
coatings, oven temperatures may range from 200°F to 450°F. At
these higher temperatures the resins and other components can
release small amounts of high molecular weight vapors. The
concentration of solvents in the oven air varies widely, but
usually lies between 200 and 2000 ppm.
Solvents used in automotive paints, enamels, and lacquers
are numerous. A list of solvents identified with surface coatings
is given in Table 3.2-2.
-21-
-------
RADEAN
CORPORATION
4.0 CARBON ADSORPTION SYSTEMS. PROCESS DESIGN
There are several means of removing organic solvent vapors
from air streams. These include incineration, catalytic incineration,
scrubbing, and carbon adsorption/incineration. The use of carbon
adsorption/incineration systems for processing air which contains
relatively low concentrations of solvents appears to be the most
feasible (CH-275, GR-193). Carbon adsorption systems are currently
in wide use in many industries for solvent removal and/or recovery.
In the past most applications were justified on an economic basis
through recovery and reuse of valuable solvents. More recently,
however, they have come into use as a means of reducing air
pollution by controlling organic materials present in gas streams
emitted to the atmosphere from many industrial operations.
As part of this study conceptual process designs were
developed for carbon adsorption/incineration systems to control
auto assembly paint plant solvent emissions A number of systems
and process variables were considered, and designs were completed
for treating paint spray booth and paint bake oven air. These
items are discussed and described below.
Section 4.1 presents general design considerations,
Section 4.2 describes the design of a spray booth control system,
and Section 4.3 describes the design of a bake oven control system.
The control system design for a typical auto assembly plant is
discussed in Section 4.4.
4.1 Design Considerations
Several types of adsorbers, a number of process variables,
and several methods of carbon regeneration were considered in
selecting and designing the system. These are described in the
following sections.
-22-
-------
RADIAN
CORPORATION
4.1.1 Types of Adsorbers
Horizontal Tanks
These types of systems are commonly used to treat
quantities of air in the range of 10,000 - 150,000 scfm (GR-193,
LE-206). For illustrative purposes, a schematic flow diagram of
a typical two-bed carbon adsorption system for organic solvent
removal and recovery is shown in Figure 4.1-1.
Each adsorber consists of a horizontal cylindrical
tank containing a horizontal bed of activated carbon. Solvent-
laden air is fed to the adsorbing bed where it flows downward
through the bed, which may be from 12 to 36 inches deep. After
passing through the adsorber, the air is exhausted to the
atmosphere. The solvent is adsorbed on the carbon until the bed
is saturated with solvent. At this point the flow of solvent-
laden air is diverted to the other bed.
Steam is then directed into the adsorber containing the
saturated bed. The direction of steam flow is normally upward
through the bed, opposite to the flow direction of the solvent-
laden air. The steam heats the carbon bed, causing the solvents
to desorb and be carried out of the adsorber with the steam. As
shown in the Figure 4.1-1, the steam is then condensed, and the
solvent (assumed to be water-insoluble) is decanted from the
water-solvent system and recovered.
Alternatively, the steam-solvent vapor stream may be
sent directly to an incinerator if the solvent is not going to be
recovered. This type of system is considered in this study as a
means of removing solvent and disposing of it.
-23-
-------
SOLVENT - FREE
EXHAUST AIR
SOLVENT
ACTIVATED CARBON
X
*4
ACTIVATED CARBON
LOW - PRESSURE STEAM
CONDENSER
SOLVENT DECANTER
„ | J ^ SOLVENT
aJ
t
WATER
LOW - PRESSURE STEAM
FIGURE 4.1-1
SOLVENT REMOVAL WITH CARBON ADSORPTION SYSTEM
-------
RADIAN
CORPORATION
Horizontal adsorbers are generally limited to sizes in
the range of 10 feet in diameter by 30 feet long. This size vessel
can be shop fabricated and shipped to the plant location.
For larger quantities of air, more adsorbers are
added to the system with one or more being regenerated at all times.
A system to treat 236,000 scfm of air and employing 10 adsorbers is
being planned (GR-193).
Canister Banks
Another type of carbon adsorption bed is the canister.
Each canister is cylindrical and consists of an outer shell of
activated carbon which may range in thickness from 0.5 to 4 in.
The direction of flow of solvent-laden air is from the outside of
the canister through the carbon bed and out through the center
of the canister.
Individual canisters as large as 14 inches in diameter
and 40 inches long are available. This type of canister contains
80-95 pounds of carbon and has the capacity to treat up to 700 scfm
of air at a velocity of 75 feet per minute across the bed. These
canisters can be mounted in banks of up to 42 in vessels having a
diameter of about 10 feet and a height of about 14 feet. Each
bank of canisters has a capacity to treat up to 30,000 scfm of
air (LE-205).
Vertical Beds
Adsorbers with horizontal beds require a considerable
land area if large quantities of air are to be processed. For this
reason, large vertical beds were considered as an alternative to
horizontal beds for this study. The carbon bed consists of a
-25-
-------
RADIAN
CORPORATION
concentric ring of carbon inside the vessel. The width of the
ring (depth of the bed) can be any size, but 12-18 inches was the
width selected in this study. A vessel of 20 ft in diameter and 50
ft high with a carbon bed having an inside diameter of 14 ft and an
outside diameter of 16 ft can process 225,000 scfm of air with a
velocity of 110 ft per minute across the bed. Thus, a system
containing 3 such adsorbers can process 450,000 scfm continuously.
Approximately 20 horizontal adsorbers would be required to process
the same amount of air.
4.1.2 Process Variables
There are a large number of process variables that should
be considered in the design of a carbon adsorption system, and
the most pertinent of these are discussed below.
Type of Solvent
Many of the solvents associated with the surface coating
industries are listed in Table 3.2-2. Included in this Table are
the normal boiling points of these compounds and the molar volume,
Vm, at the normal boiling point. Compounds of larger molar volume
tend to displace those of smaller volume during adsorption on
activated carbon. The certainty of displacement increases as the
difference in Vm increases (MS-004), so that solvents with small
Vm become increasingly difficult to control with activated carbon
adsorption systems. Solvents with Vm larger than 80 cm3/mol offer
no real problem, at least in the adsorption phase of the cycle,
while problems with those with Vm less than 80 cm3/mol are vari-
able. The problems of desorption increase significantly with
solvents having boiling points above 400°F.
-26-
-------
RADIAN
CORPORATION
In general, the surface coating industries use hydrocarbons
(aromatic, naphthenic, and aliphatic), alcohols, ketones, ethers,
glycol ethers, esters, and nitroparaffins. Solvents are used in
blends of two to twelve to obtain the desired viscosity, solvency
power, drying rate and compatibility with the drying oils and
resins. Most frequently, the blends consist of two to five solvents
(EN-321, MS-001) . The solvents appearing most often in the blends
are toluene and xylene.
When the coatings are baked or cured in the baking ovens
at temperatures up to 450°F, small amounts of high boiling resins
and decomposition products can be volatilized. Unless these are
removed prior to entering the adsorption bed, they will be adsorbed.
Due to their relatively high molecular weights and boiling points,
these compounds may be difficult to desorb using low pressure
saturated steam. Superheated steam or hot inert gas may be required.
Some solvents are unstable when in contact with carbon.
For example, 4-methyl-2-pentanone can polymerize, thus shortening
the carbon bed life. Some ketones decompose and this can create
hotspots and subsequent carbon bed fires at high ketone concentra-
tions and/or if proper preventive procedures are not followed.
Solvent Concentration
The solvent concentration range of 1-500 vppm is impor-
tant for solvent emissions considerations. This concentration
range has not been seriously considered in previous solvent
removal technology development. Nonregenerative air purification
systems may be more practical and economical for treating gases in
the concentration range below 10 vppm than regenerative types.
Above this range the cost of frequent carbon replacement becomes
prohibitive, and regenerative systems are favored.
-27-
-------
RADIAN
CORPORATION
In general, the efficiency of the carbon adsorption
system increases with increasing solvent concentration, since
both the absolute and working capacities are greater at higher
concentrations. Studies indicate that the economic breakeven
concentration for recovering a single, water-insoluble solvent
averages about 700 vppm. Around this point the value of the
recovered solvent is just about equal to the cost incurred in
recovering it (MS-004).
At high solvent concentrations, the heat of adsorption
can be high enough to cause a significant rise in bed temperature.
This can be sufficient to reduce the adsorptive capacity of the
bed and may even cause vapor decomposition. At the relatively
low concentrations encountered in most auto plant paint opera-
tions, however, the temperature rise should not be great enough
to cause any problems.
Multiple Solvents
Air streams containing multiple solvents are treated
in many commercial carbon adsorption installations (VU-003, GR-193).
The use of mixed solvents does, however, reduce the adsorptive
capacity of carbon beds (MS-004) for the individual solvents. Less
carbon is required when removing a single solvent at a given con-
centration than is needed to remove a mixture of solvents with the
same combined concentration.
When air containing a mixture of solvents is passed
through a bed of carbon, the most strongly adsorbed vapors are
separated out in the part of the bed closest to the air inlet.
The other solvents are adsorbed in successive bed segments in
order of decreasing adsorbability. As the adsorption is continued,
-28-
-------
RADIAN
CORPORATION
the more-strongly adsorbed solvents displace the less-strongly
adsorbed ones in the adsorption zone. Depending on the extent of
the differences in adsorptivities, either complete displacement or
partial displacement accompanied by some coadsorption may occur in
the adsorption zone.
The simplest design procedure for processes treating
multiple solvents is to assume that each solvent is adsorbed
in separate successive beds. This method is conservative and
usually results in a bed of excessive depth because of two factors.
Coadsorption is generally present in the saturated and adsorption
zones. Adsorptive capacities in some zones may increase due to
desorption in other zones, which produces an effective rise in
vapor concentrations.
Activated Carbon
Activated carbon is the adsorbent most suitable for
removing organic vapors (DA-069, MS-004). Carbon adsorbs
substantially all the organic vapors from the air at ambient
temperatures regardless of variation in concentration. Being
generally a non-polar material, carbon selectively adsorbs non-
polar vapors. Organic vapors are adsorbed in preference to
oxygen, nitrogen, and of particular interest, water. It is
generally accepted that compounds with a molecular weight of at
least 45 or with a boiling point over 0°C will be well adsorbed
on activated carbon (DA-069, GR-193).
Typical granulated activated carbons have high pore
volumes of 0.80-1.20 cm3 per gram and surface areas in the range
of 600-1600 m2 per gram of carbon. The carbons must be strong
enough to withstand shear, compressive, and impactive forces
encountered in handling and use. Carbons are manufactured from
sources such as bituminous coal and coconut shells.
-29-
-------
RADIAN
CORPORATION
The selection of a carbon adsorbent requires considera-
tion of variables such as types of solvents, solvent concentration,
operating conditions (temperature, pressure, humidity, etc.).
required removal efficiency, reactivity of solvents on carbon,
and contact time.
The ultimate capacity of activated carbon for hydrocarbon
vapors is quite high and may exceed .40 cm3 liquid per gram of carbon
at ambient temperatures. The actual working capacity, however, is
considerably less than the ultimate capacity and declines with
decreasing solvent concentration. The working capacity is the
amount of solvent that is repeatedly adsorbed and desorbed during
each cycle. It is not economically practical to remove all of
the solvent from the adsorbent on each cycle. In general, a carbon
can typically be expected to have a working capacity of 5-20 per-
cent of its own weight (MS-004, LE-207, MA-491).
In shallow beds where high efficiencies are required,
granulated carbon particles as small as 12-28 mesh size may be used.
In deep beds, however, coarser carbons of 4-6, 6-8, or 4-10 mesh
size are generally employed to minimize pressure drop, even though
a lengthened adsorption zone is required.
Bed Depth
The bed depth depends on the inlet and outlet solvent
concentrations, length of adsorption zone, working capacity of
the carbon, temperature, number of beds and allowable pressure
drop. Bed lengths in solvent recovery/removal systems are
normally in the range of 8 to 36 inches. Deeper beds provide
more capacity, fewer regenerations, and longer cycle lengths than
shallower beds. This advantage must be balanced against the
higher pressure drop of deeper beds with correspondingly larger
blower and power requirements.
-30-
-------
RADIAN
CORPORATION
Gas Velocity
The frequency of the adsorption/desorption cycle, the
bed dimensions, the solvent removal efficiency, and the resistance
to flow must be considered in selecting a gas velocity. Air
velocities commonly used in carbon adsorption systems range from
20 to 110 feet per minute 01S-004, LE-205). Velocities above
110 feet per minute cause high resistance to flow, operational
problems with valves, extended adsorption zones, and physical
attrition of the carbon. The relatively low allowable or practical
velocity of the gas across the bed leads to large bed areas and
large adsorbers.
Temperature
With most types of solvents and ranges of concentration,
the air entering the carbon adsorption bed should not exceed 100°F
(GR-193, VU-003). Adsorption should be carried out at a tempera-
ture as low as practical because the capacity decreases signifi-
cantly with relatively small increases in temperature. If the
solvent-laden air temperature is above 120°F, it is generally more
economical to cool the air than to operate at reduced capacity
(LE-205, MS-004). At 120°F, the capacity of carbon is about 75%
of the capacity at 77°F. Where air cooling is required the adsorber
beds are generally operated at 95-105°F (LE-207, MS-004).
Aging of the carbon bed normally occurs through carbon
oxidation. The oxidation rate increases if the adsorbers are oper-
ated continuously at temperatures much above 120°F (VU-003).
-31-
-------
RADIAN
CORPORATION
Humidity
Large amounts of water vapor (a relative humidity of
80% or more) in the solvent-laden air stream interfere with the
operation of the carbon adsorption system. After steam regenera-
tion and cooling the bed still contains a considerable amount of
moisture. The heat produced by adsorption of solvent on the carbon
causes vaporization-of the remaining water from the bed. If the
relative humidity of the solvent-laden air stream is high, the
water in the bed will not vaporize and water will occupy some
sites on the carbon that are normally available for the adsorption
solvent. Thus, the working capacity of the bed is reduced.
Particulates
Particulates from paint spray booths consist almost
entirely of dried paint particles which escaped capture in the
water wash system. Since these materials tend to form coatings
on surfaces that they contact, they interfere with the operation
of carbon adsorbers. The particulate matter may be deposited
on the surfaces of coolers and blowers, causing a decline in
operating efficiencies. Periodic cleaning of this equipment is
then required. Particulates also plug the carbon adsorption beds,
requiring costly replacement of the bed.
The particulate matter may be removed from the air stream
with fabric filters made of organic and/or glass fibers. An inter-
mediate or roughing filter may in itself be sufficient to protect
the bed. In some cases the intermediate filter should be followed
by a high-efficiency filter. Fabric filters are available in many
efficiencies, but the cost and resistance to flow increase as the
-32-
-------
RADIAN
CORPORATION
efficiency improves. The initial resistance of intermediate
filters is usually between 0.2 and 0.6 inches of water. The
resistance rises to 1.0 inch of water as the filter becomes
clogged with particulate matter. It is usually replaced at
this point.
In a pilot plant study, Chass, et al. (CH-275) found
that carbon beds could be adequately protected from both enamel
and lacquer paint spray particles by selection of a proper filter
system.
4.1.3 Carbon Regeneration
When the carbon bed becomes saturated with adsorbed
solvent and the solvent begins to pass through the bed without
being completely adsorbed, the carbon is regenerated. This is
accomplished by heating the bed to desorb the solvent and sweep-
ing the desorbed solvent vapor from the bed using a clean (solvent-
free) gas stream. Steam is generally chosen as the regeneration
gas. Hot air or inest gas are also used. These are discussed
below.
Steam Regeneration
Most regenerative systems in current use are employed in
solvent recovery operations. The most common regenerating agent
for this application is steam. In general, steam is a relatively
economical means of transferring heat into the bed of charcoal.
Enough heat must be supplied to bring the temperature of the
carbon up to the desired desorption temperature and to provide
the heat of desorption. Steam not only has a very high specific
-33-
-------
RADIAN
CORPORATOON
heat, but the heat of condensation of the steam is utilized as it
condenses on the adsorbent. It is desirable to operate as close
to the dew point of the steam as possible to fully utilize the
heat of condensation. In some cases, however, superheated steam
is required in order to desorb high boiling solvents.
The amount of steam required for regeneration depends
on the operating capacity of the carbon, the cycle length, and
the solvent concentration in the solvent-laden air. As desorption
of the adsorbed material approaches completion, the steam require-
ments increase greatly. In order to minimize the steam/solvent
ratio, the extent of desorption (working capacity) must be
optimized.
As the solvent concentration in the incoming air decreases,
the required cycle time and amount of regenerating agent increases.
Steam/solvent ratios found in the literature vary widely. A sum-
mary of some of these ratios is given in Table 4.1-1. Two to four
pounds of steam per pound of desorbed solvent are considered to
be optimum ratios (LE-206). The use of 9-12 pounds of steam per
pound of solvent is reported for a styrene solvent recovery system
(GR-193), but styrene concentrations are not specified. For
solvent recovery systems Vulcan Cincinnati literature (VU-003)
mentions steam/solvent ratios of 2-10 pounds steam per pound of
recovered solvent. In an article describing a system which uses
flue gas to regenerate the carbon adsorbent, Mattia states that
reducing solvent concentration to 200 vppm can result in steam
requirements in excess of 30 pounds per pound of recovered solvent
(MA-050). Cannon (CA-288) suggests a steam/solvent ratio of 5.3
pounds per pound to regenerate carbon exposed to air containing
100 vppm of solvent.
-34-
-------
TABLE 4.1-1
STEAM/SOLVENT RATIOS FOR CARBON ADSORPTION
Solvent
Concentration Steam/Solvent
Solvent vppm lb/ lb Reference
Varied Varied 2-4 LE-206
Styrene - 9-12 GR-193
Varied Varied 2-10 VU-003
200 30 MA-050
100 5.3 CA-288
Toluene 150 13 CH-275
10 80-400 MS-004
-------
RADIAN
CORPORATION
Chass, et al., (CH-275) used an experimentally derived
value of 13 pounds steam per pound of recovered toluene in their
estimation of the cost of a solvent vapor removal system. The
incoming solvent-laden air contained 150 vppm toluene, and the
carbon had a working capacity of 8.5 pounds of toluene per pound
of carbon.
Some high-boiling compounds may require the use of
superheated steam for adequate desorption. The use of super-
heated steam results in lower steam requirements per pound of
solvent due to the higher steam temperature. This is shown in
results obtained by MSA Research Corporation (MS-004) in an
investigation of the performance of 260°F steam. Superheated
steam may also be intermittently used to remove high-boiling
compounds which are present in low concentrations in the polluted
air and are not completely removed in regenerations with low-
pressure saturated steam.
An additional advantage of using steam as the regen-
erating medium is that the possibility of explosive conditions
during the regeneration cycle is prevented.
Air or Inert Gas Regeneration
In some applications where solvent concentrations are
low and, in addition, the solvents have little or no recovery
value, steam may not be the best regenerating agent. If the
quantity of required steam is quite large relative to the amount
of adsorbed solvent, it may be more economical in some cases to
regenerate with a noncondensable gas such as air or inert flue
gas.
Regeneration with hot air does present some problems.
The specific heat of hot air is quite low, and large volumes of
-36-
-------
RADIAN
CORPORATION
hot air must be employed to desorb the solvent. Hot air regenera-
tion can create conditions in which there is an ample amount of
oxygen present for explosion at elevated temperatures. Furthermore,
hot air causes additional oxidation of the carbon, thereby reducing
its effective life.
Regeneration with flue gas avoids the problem of potential
explosion or combustion in the adsorber. Mattia (MA-050) presented
some preliminary studies which indicated that a solvent removal
process using flue gas regeneration with subsequent incineration
(ZORBCIN Process) is potentially more economical than one employ-
ing steam for regeneration with subsequent recovery of the solvent.
The complexity of the ZORBCIN system is somewhat greater than the
steam regeneration system.
-37-
-------
RADIAN
CORPORATION
4.2 Paint Spray Booth Solvent Emission Control
Paint spray booth effluent control systems were defined
for a base study case. A conceptual process design was developed
for this system and is discussed in this section. Major equipment
items were selected and sized, and these are described. Varia-
tions of the base case design were also investigated.
4.2.1 Description of System
A typical automotive spray paint booth was defined as
one which is exhausting 450,000 scfm of solvent-laden air to the
atmosphere. The air was assumed to have passed through a water
curtain to remove paint particulates, and to be available at
100°F. As mentioned in a previous section of this report, the
number of solvents contained in a paint coating may be as high
as twelve. It is unnecessary for conceptual design purposes to
specify the number, identity, and concentrations of solvents.
Toluene was selected as a typical solvent, and a concentration of
50 vppm in the spray booth effluent air was chosen.
A working capacity for the carbon adsorbent was defined
as 7 percent by weight (MA-491). The carbon working capacity for
toluene is relatively low compared to many other solvents (MA-491).
The size selected for the carbon was 6-8 mesh, and the bulk
density of the carbon was assumed to be 28 pounds per cubic foot.
Toluene removal efficiency of the carbon bed was assumed
to be 90%. Steam was chosen as the regenerating agent for three
reasons. Equipment using this regeneration medium is relatively
simple. Processes using steam regenerant are in commercial opera-
tion. In addition, the danger of explosion is virtually eliminated,
even in the event of equipment malfunction.
-38-
-------
RADIAN
CORPORATION
A carbon life of 4 years was assumed with a stream time
of 5500 hours per year. After this time, a fresh charge of carbon
is loaded into the adsorbers.
The bases for Case III, the base case, and other cases
are summarized in Table 4.2-1.
4.2.2 Equipment Description
A schematic flow diagram for a carbon adsorption system
to treat 450,000 scfm of solvent-laden air (SLA) is shown in
Figure 4.2-1. Three adsorption vessels are used, each with a
concentric or annular bed of carbon. At any given time two of
the beds are adsorbing and one is being regenerated. The solvent-
laden air enters from the paint spray booth, is filtered to
remove particulate matter, and continues on into the SLA blower.
This blower provides enough head to move the air through the
carbon bed and out the top of the adsorption vessels. In the
adsorption vessels, the solvent-laden air enters the outer portion
of the vessel, flows horizontally across the carbon bed where
solvent is adsorbed, and passes up through the center of the bed
to be exhausted to the atmosphere.
During regeneration, steam (15 psig, saturated) is
passed through the bed in the direction opposite to that followed
by the air. As it passes across the bed, the solvent is desorbed
and is carried out of the bed with the steam. After mixing with
preheated combustion air and passing through a booster fan which
provides enough head to get the vapor into the incinerator, the
mixture is incinerated in the incinerator/boiler. There is
enough solvent in the vapor (under average desorption conditions)
to maintain combustion, but some fuel (gas or oil) is added for
control purposes as well as to provide steam until the bed is
heated and solvent begins to desorb.
-39-
-------
TABLE A. 2-1
BASES FOR DESIGN AND CASE STUDIES
CASE
Feed
Air Source
Air Temperature, "F
Solvent
Solvent Concn., ppm
Air Flow Rate, 10' scfm
Booth
100
SO
60
IIA
Booth
100
IIB
III
Case Caae\
Booth J
IV
Booth
100
Booth
100
Booth
100
Booth
100
50
2S0
50
250
50
450
50
900
50
450
VI
Booth
100
50
450
VII
Booth
100
Toluene
150
450
VIII
Booth
100
300
450
IX
300
10
XI
Oven
375
fase Case t
Oven /
XII
Oven
375
300
20
Oven
375
300
50
Oven
375
300
250
I
¦f-
O
I
Adsorption
Bed Depth, Inches
Carbon Size, Mesh
Max Air Velocity
Across Bed, ft/mln
Carbon Operating
Capacity, wt X
Solvent Removed, X
Carbon Life, years
Regeneration Steam
Steam Pressure, pslg
Steam Required lb/
lb solvent
Cycle Length, hr.
Adsorbing
Regenerating
Cooling
Stream Time
Hours/day
Days/week
Weeks/year
Hours/year
12
12
12
12 12 12 12
12 18 18
6-8 -
18 18
28
28
14
24
24
36
60
6
5
6
6
5
6
6
8
2
1
6
1
6
6
7
90
4
15
10
9
4
0.5
20
6
46
5500
6 8 7 5
3 0 5 0
0.4 2.5
7.5
5 0
2 5
7 5
3.0
0.8
18
110 110 110 110 110 80 50 110 110 110 110 110 110
8.0
2 0
0.7
-------
VENT
VENT
VENT
STEAM
ADS
OES
AOS
FRESH AIR -Q-Q
SOLVENT LADEN AIR -o—
FROM PAINT SPRAY
BOOTHS
FLUE GAS
FUEL
INCINERATOR
- BOILER
AIR
BOOSTER FAN
BOILER FEED WATER
TO HEAT RECOVERY
FIGURE 4.2-1
CARBON ADSORPTION SYSTEM FOR PAINT SPRAY BOOTH SOLVENT EMISSION CONTROL
CASE III (BASE CASE): 450,000 SCFM OF SOLVENT - LADEN AIR
-------
RADIAN
CORPORATION
Enough steam is generated in the boiler by the inciner-
ation to provide the amount required for regeneration. The flue
gas from the incinerator is still hot enough to provide heat for:
1) preheating the air required for combustion, and 2) production
of additional steam for use in the plant or preheating of the air
going to the baking ovens.
The major pieces of equipment are listed in Table 4.2-2.
The major equipment items, sized for the Base Case III, are
described below. Costs of the major pieces of equipment are
included. The bases for these costs are given in Section 5.1
Adsorbers
The shell for each adsorber is a vertical cylindrical
vessel. For the Base Case III (450,000 scfm) each adsorber is
50 feet high and 20 feet in diameter. The carbon is enclosed in
a concentric or annular bed 12 inches thick. Some solvents are
themselves corrosive to mild steel and others decompose during
adsorption or regeneration to form corrosive compounds. For this
reason the shell is clad with stainless steel and the adsorber
internals are constructed of stainless steel. The bed itself is
45 feet high, and the face area of the bed is 2050 square feet.
SLA Blower
The solvent-laden air blower has a capacity of 485,000
cfm of air at 100°F (450,000 scfm). A head of 17 inches of
water is sufficient to move the air through the system, including
the 12 inches of activated carbon. The blower is sized to oper-
ate normally at 807„ of its rated capacity with an overall efficiency
of 65%. This requires a 2500 horsepower motor. The blower itself
is constructed of carbon steel.
-42-
-------
RADIAN
CORPORATION
TABLE 4.2-2
CARBON ADSORPTION SYSTEM FOR SPRAY BOOTH EMISSION CONTROL
CASE III (BASE CASE): EQUIPMENT LIST
1. Adsorbers Number Required: 3
Size: 20 ft. diameter, 50 ft. high
Carbon Bed: Concentric, 12 inches
thick, 14 ft. i.d., 16
ft. o.d.
Material: Shell - 316 SS clad
Internals - 316 SS
Cost: $533,000 each (Installed, 1975)
$1,599,000 for 3 adsorbers (excl.
carbon)
2. SLA Blower Number Required: 1
Capacity: 485,000 cfm
Discharge Head: 17 inches of water
Material: Carbon steel
Size: 2500 bhp (65% efficiency @ 807o
capacity)
Cost: $151,000 (installed 1975)
3. Boiler/Incinerator Number Required: 1
Capacity: 12,000 pounds of steam per
hour
Steam Condition: 15 psig, saturated
Cost: $33,000 (installed, 1975)
4. Control Valves Number Required: 6
Size: 48 inch
Type: Butterfly
Material: Carbon steel
Cost: $20,000 each (installed, 1975)
$120,000 for 6 valves
-43-
-------
RADIAN
CORPORATION
TABLE 4.2-2 (Continued)
Number Required: 1
Area: 150 square feet
Material: Carbon steel/stainless steel
Cost: $24,000 (installed, 1975)
Capacity: 485,000 cfm
Type: Disposable fiberglass bag filters
Cost: $50,000 (installed, 1975)
Number Required: 1
Capacity: 2200 scfm
Material: Carbon steel
Size: 20 bhp
Cost: $6000 (installed, 1975)
8. Incincerator Booster Number Required: 1
Capacity: 5500 cfm
Material: Carbon steel
Size: 20 bhp
Cost: $6000 (installed, 1975)
5. Air Preheater
6. Air Filters
7. Fresh Air Blower
-44-
-------
RADIAN
CORPORATION
Boiler/Incinerator
The boiler/incinerator is assumed to be a "package"
type boiler with alterations to permit incineration/combustion
of the regeneration stream of solvent and steam and to allow
auxiliary firing with gas or oil. The boiler was sized for twice
the average required steam rate.
Control Valves
Six 48-inch control valves are required for the air inlet
and outlet lines of the three adsorbers. These are carbon steel
butterfly-type valves with Teflon seals. They are required to
close and remain virtually leak free against a pressure of 30
inches of water. Minimal leakage is required in order to insure
that solvent-laden air will not leak into an adsorber during the
regeneration cycle. Other valves were not itemized because they
are of nominal size and cost.
Air Preheater
A small shell and tube exchanger is required to preheat
the combustion air that is mixed with the regeneration stream of
steam and solvent. The air is preheated to 220°F to prevent
cooling and condensation of the steam in the regeneration stream.
The tubes are constructed of stainless steel to prevent damage
from any corrosive components that may be present in the regenera-
tion gas. An exchanger with an area of 150 square feet is required.
Air Filters
The solvent-laden air from the paint spray booth will
contain fine, dry particles of paint pigment. The actual partic-
ulate loading in the air is not known and will depend on the
efficiency of the overspray removal system (water curtain, baffles,
-45-
-------
RADIAN
CORPORATION
etc.) and the degree of filtration already present in the assembly
plant. It is necessary to keep the carbon bed as free of particu-
lates as possible. There are many types of dry, disposable filter
mediums available including felt fabric and glass fabric bags,
envelopes, pads, etc. Disposable fiberglass bag filters were
selected because of their high efficiency. It was assumed that
replacement of these filters would be reqired on a monthly basis
A total of 4-85,000 cfm of air at 100°F must be filtered.
Fresh Air and Incinerator Blowers
Two additional small blowers are required in the process.
A fresh air blower is required to blow fresh air through the
adsorber after regeneration in order to cool the bed to approxi-
mately 100-110°F before the adsorber is placed back in service.
The fresh air blower must have a capacity of about 2200 scfm and
must provide a discharge head of about 6-8 inches of water. A
10 horsepower motor was selected to drive the blower. The blower
itself is constructed of carbon steel.
The incinerator blower is needed in order to boost the
pressure of the stream of steam, solvent vapor, and air to a
pressure sufficiently high to move the stream through the burner
and incinerator. A discharge head of 10-15 inches of water is
required. An average of 5500 cfm of gas at 220°F is handled, and
a 20 horsepower motor was selected in order to give a reasonable
degree of flexibility in handling surges of gas. This blower is
also made of carbon steel.
4.2.3 Case Studies
A number of case studies were made. The effect of
several major variables on the arrangement, size, and type of
process equipment required to treat solvent-laden air from paint
spray booths was investigated in Cases I through VIII. The
-46-
-------
RADIAN
CORPORATION
principal variables examined were system size (Cases I-IV),
allowable air velocity through the carbon bed (Cases V and VI),
and solvent concentration in the solvent-laden air (Cases VII
and VIII). The bases for these cases are tabulated in Table 4.2-1.
4.3 Paint Oven Solvent Emission Control
A base case was defined for a carbon adsorption/incinera-
tion system to remove solvent vapors from paint oven effluent air.
The conceptual process design for the base case as well as other
cases is discussed in this section. The size and type of required
equipment is described.
4.3.1 Description of System
A base case (Case XI, in Table 4.2-1) was defined for
a system to treat 50,000 scfm of solvent-laden air from an auto
assembly plant paint oven. The solvent, toluene, is present in
the air at a concentration of 300 vppm. A schematic flow diagram
for the base case process is shown in Figure 4.3-1. In its basic
concept and design, the process is similar to the one developed
to treat paint spray booth air. There are some important differ-
ences, however.
The air from the paint oven is hot when exhausted. For
design purposes, it is assumed to be at 375°F when delivered to
the adsorption system. Cooling is required to reduce the tempera-
ture of the air to 100°F before it enters the carbon adsorption
bed, since the bed will not function efficiently at temperatures
much above 100°F.
Small amounts of high boiling materials such as resins,
polymers, or plasticizers are generally present in the air from
paint ovens. If these components are adsorbed on the carbon,
-47-
-------
CLEAN AIR TO
ATMOSPHERE
AIR COOLER
CW
ADSORBER
AIR
FILTER
ADSORBER
AIR FROM
PAINT
OVEN
FLUE GAS TO
HEAT RECOVERY
CW
AIR
BLOWER
ADSORBER
AIR
AIR COOLER
STEAM
INCINERATOR
-BOILER
BOILER
FEED
WATER
AUXILIARY FUEL
FIGURE 4.3-1
CARBON ADSORPTION SYSTEM FOR PAINT OVEN SOLVENT EMISSION CONTROL
BASE CASE XI: 50,000 SCFM
-------
RADIAN
CORPORATION
removal by steam regeneration is difficult, and the life of the
carbon is, therefore, reduced. However, during cooling from
375°F to 100°F in the air coolers, it is very possible that most
of the high-boiling components present in the vapor stream will
condense and be deposited on the air side of the cooling surface.
This will protect the carbon bed to a large extent, but the air
coolers will have to be periodically cleaned to remove the con-
densed heavy materials. For this reason, two exchangers are
required. One will be in service while the other is being cleaned.
The cooling medium is water available at 75°F and returning to
the cooling tower at 90°F.
After being cooled, the solvent-laden air is filtered
to remove particulates. It then flows to an air blower which
provides the head required to move the air through the entire
solvent removal system. The air passes from the blower into the
horizontal adsorbers. For this 3-adsorber system, two of the
beds are always adsorbing, while the other is undergoing regenera-
tion. The air passes through the horizontal carbon beds and is
exhausted to the atmosphere after 90% of the solvent vapors have
been removed.
Regeneration is accomplished with steam in the same
manner as described in the previous section, and the flow of the
regeneration vapor stream is also the same, i.e., from the adsorber
to the incinerator/boiler.
The life of the carbon bed was assumed to be only two
years due to possible buildup of high boiling materials. After
two years a fresh charge of carbon will be required.
The bases for the oven base case, Case XI, and other
design cases are given in Table 4.2-1.
-49-
-------
RADIAN
CORPORATION
4.3.2 Equipment Description
The major pieces of equipment required for the base
case system were determined and sized. They are shown schemati-
cally in Figure 4.3-1, listed in Table 4.3-1, and described in
the following section. Except for size, the filters, blowers,
boiler-incinerator, control valves, and air preheater are the
same as those selected for the spray booth control system. These
items were described in Section 4.2.2 and, therefore, will not
be discussed here.
Adsorbers
Three adsorbers are specified in the base case design.
They consist of cylindrical tanks erected horizontally, and each
contains a horizontal bed of granulated carbon of 6-8 mesh size.
The beds are 18 inches deep, and each has a face area of 250
square feet and contains 10,500 pounds of carbon. The vessel
shells are rated at 50 psig and are clad in 316 stainless steel
to prevent corrosion. The vessel internals are constructed
entirely of 316 stainless steel. Each tank is 10 feet in diameter
and 25 feet long.
The solvent-laden air enters the top of each tank and
flows downward through the bed of carbon. During regeneration
the steam flows in the opposite direction, i.e., in at the bottom
of the tank, upward through the bed, and out the top of the
adsorber.
Air Coolers
Floating-head shell and tube exchangers were selected
as a means of cooling the oven air from 375 °F to 100°F. There
will probably be some condensation in the exchanger of high
-50-
-------
RADIAN
CORPORATION
TABLE 4.3-1
CARBON ADSORPTION SYSTEM FOR PAINT OVEN EMISSION CONTROL
CASE XI (BASE CASE): EQUIPMENT LIST
1. Adsorbers Number Required: 3
Size: 10 feet diameter,
25 feet length (horizontal)
Carbon Bed: Horizontal, 18 inches deep
Material: Shell - 316 SS clad
Internals - 316 SS
Cost: $60,600 each (installed, 1975)
$181,000 total (excl. carbon)
2. Air Coolers Number Required: 2
Area: 7400 square feet per exchanger
Type: Shell and tube, floating head
Material: CS/CS
Cost: $158,900 each (installed, 1975)
$317,800 total
3. SLA Blower Number Required: 1
Capacity: 54,000 cfm
Head: 30 inches of water
Material: Carbon steel
Size: 500 bhp
Cost: $74,800 (installed, 1975)
4. Boiler/Incinerator Number Required: 1
Capacity: 5000 pounds of steam per hour
Steam: 15 psig, saturated
Cost: $14,300 (installed, 1975)
-51-
-------
RADIAN
CORPORATION
TABLE 4.3-1 (Cont.)
5. Control Valves
6. Air Preheater
7. Air Filters
8. Fresh Air Blower
9. Incincerator Fan
Number Required: 6
Size: 3-30"
3-24"
Type: Butterfly
Cost: 30" - $12,700 each
24" - $10,000 each
$68,100 total (installed, 1975)
Number Required: 1
Area: 50 square feet
Material: CS/SS
Cost: $9,300 (installed, 1975)
Capacity: 80,000 cfm
Type: Disposable fiberglass bag filters
Cost: $16,000 (installed, 1975)
Number Required: 1
Capacity: 550 scfm
Material: Carbon steel
Size: 2 bhp
Cost: $1800 (installed, 1975)
Number Required: 1
Capacity: 2100 cfm
Material: Carbon steel
Size: 10 bhp
Cost: $5300 (installed, 1975)
-52-
-------
RADIAN
CORPORATION
boiling materials present in the air. The exchangers will
require periodic cleaning to remove this material. For this
reason two exchangers are required. One will always be in
service. Air will be routed through the tubes to facilitate
cleaning procedures, since deposits will form in the tubes.
Cooling water, available at 75 °F and returned to the
cooling tower at 90°F, will be used as the cooling medium. Each
exchanger is constructed of carbon steel and contains 7400 square
feet of heat exchange surface.
4.3.3 Other Case Studies
The only parameter investigated for its effect on the
design and cost of an adsorption system for paint oven emissions
was the volume of oven air to be treated. In addition to the
base case (Case XI), which was sized for 50,000 scfm of solvent -
laden air, three other systems were designed for 10,000, 20,000,
and 250,000 scfm of air. The bases for these systems, Cases
IX, X, and XII, are given in Table 4.2-1
4.4 Emission Control System for a Typical Auto Assembly
Plant
A typical auto assembly plant paint operation was
defined. The plant contains two assembly lines (lines A and B),
and each of the assembly lines incorporates paint spray booths
and paint bake ovens. While as many as 10 spray booths and 10
ovens may be included in each line, it was assumed that solvent
emission controls are required on only two booths and four ovens
from each line. For compliance with regulations it is required
that the emissions from each paint spray booth be reduced to 3000
pounds per day, and that the emissions from the paint ovens be
-53-
-------
RADIAN
CORPORATION
reduced by 857». This requires total reductions of 6400 and 1150
pounds of solvent per day from the paint spray booths and bake
ovens, respectively.
The assumed solvent distribution in the air from the
various sources is shown in Table 4.4-1. The solvent concentra-
tions are 50 vppm in air from the spray booths and 300 vppm in
air from the paint ovens. Air from the spray booths is available
at 100°F from the plant, while the solvent-laden air from the
ovens is emitted at 375°F.
4.4.1 System Description
Two large adsorption systems are required to treat the
air from the paint spray booths. It was assumed that only enough
air would be treated to reduce the emissions rates to allowable
levels. These levels were 3000 pounds per day from each spray booth
and 85% reduction in the emissions from each oven. A 90 percent
solvent removal efficiency is specified. The minimum amount of
air to be treated is 320,000 scfm and 170,000 scfm from the spray
booths of assembly lines A and B, respectively. To allow for some
flexibility, the systems were sized for 350,000 scfm and 200,000 scfm.
The quantity of air to be treated from the paint ovens
is considerably less than that from the spray booths. Again, it
was assumed that only enough air would be treated to effect the
required reductions. The amount of air from the paint ovens that
requires treatment is 6300 scfm from line A and 8200 scfm from
line B. Both adsorption systems were sized to treat 10,000 scfm.
It is recognized that multiple solvents are present in
spray booth and paint oven air, and that the presence of several
solvents instead of one could require a somewhat deeper carbon
bed to effect proper removal. For design purposes in this case,
however, toluene was selected as a representative solvent.
-54-
-------
RADIAN
CORPORATION
TABLE 4.4-1
PAINT PLANT OPERATIONS OF A TYPICAL AUTO ASSEMBLY PLANT
SUMMARY OF SOLVENT EMISSIONS
Emission
Source
Assembly
Line
Current
Solvent
Emissions,
lb/day
Proposed
Solvent
Emission
Limits,
lb/day
Required
Reduction.
lb/day
Spray Booth
1
A
5,500
3,000
2,500
Spray Booth
2
A
4,700
3,000
1,700
Spray Booth
3
B
4,900
3,000
1,900
Spray Booth
4
B
3,300
3,000
300
18,400
12,000
6,400
Oven
1
A
176
26
150
Oven
2
A
118
18
100
Oven
3
A
176
76
150
Oven
4
A
118
18
100
Oven
5
B
235
35
200
Oven
6
B
118
18
100
Oven
7
B
294
44
250
Oven
8
B
118
18
100
1,353
203
1,150
-55-
-------
RADIAN
CORPORATION
4.4.2 Description of Equipment
The major equipment items required in the four separate
systems are listed in Table 4.4-2.
Adsorbers
The spray booth solvent adsorption systems each consist
of three vertical adsorbers, all of which are 20 feet in diameter.
The process flow is the same as shown in Figure 4.2-1. The
adsorbers treating air from the spray booths in assembly line A
are 40 feet high, while those removing solvent from assembly line
B spray booths are 30 feet high. Each adsorber contains a con-
centric or annular bed of carbon which is 12 inches thick. In
each system, two adsorbers are adsorbing and one is being regen-
erated at any given time. The adsorbers are internally clad
with 316 stainless steel, and the internals are entirely construc-
ted of 316 stainless steel.
The operating conditions are shown in Table 4.4-3.
With the exception of the normal air velocity across the bed,
the operating conditions are the same as those of Case IIB shown
in Table 4.2-1.
Two horizontal adsorbers, 5 feet in diameter and 18
feet long, are used in each of the oven emission control systems.
These vessels are also clad in 316 stainless steel and the vessel
internals are made of 316 stainless steel. The carbon beds are
18 inches deep. In each system one bed is adsorbing, and one is
undergoing regeneration at any time.
-56-
-------
TABLE 4.4-2
Air Source
SOLVENT REMOVAL SYSTEM FOR A TYPICAL AUTO ASSEMBLY PLANT
MAJOR EQUIPMENT ITEMS
Assembly Line
No. of Sources
System ID
Raced System Capacity, scfm
Paint Spray Booths
A
2
A1
350,000
B
2
Bl
200,000
Paint Bake Ovens
A
4
A2
10,000
B
4
B2
10,000
Adsorbers
Number 3 3
Type Vertical Vertical
Size Feet (Diameter X Length or
Height) 20 x 40 20 x 30
Horizontal Horizontal
6 x 18 6 x 18
Air Coolers
Number
Area, Ft2/Exchanger
2
1,500
2
1,500
SLA Blower
Number
bhp
1
2,000
1
1,150
1
200
1
200
Incinerator-Boiler
Number
Capacity, lb steam/hour
1
11,000
1
5,000
1
1,250
1
1,250
Combustion Air Preheater
Number
Area, Ft1
1
105
1
80
1
15
L
15
Control Valves
Type
Number
Size, inches in diameter
Butterfly Butterfly
6 6
48 48
Butterfly Butterfly
4 4
30 & 24 30 & 24
-57-
-------
TABLE 4 4-3
SOLVENT CONTROL SYSTEM FOR TYPICAL AU1Q ASSEMBLY PLANT
OPERATING CONDITIONS
Air Source
Assembly Line
Air Temperature, °F
Solvent
Solvent Concentration, vppm
Design Air Flow Rate, scfm
Normal Air Flow Rate, scfm
Paint Spray Booths
A
100
Toluene
50
350,000
320,000
B
100
Toluene
50
200,000
170,000
A
375
Toluene
300
10,000
6,300
Paint Ovens
B
375
Toluene
300
10,000
8,200
Adsorption
Bed Depth, inches
Carbon Size, mesh
Air Velocity Across Bed
Maximum, ft/min
Normal, ft/min
Carbon Operating Capacity, wt %
Solvent Removal Efficiency, %
Carbon Life, years
Operating Temperature, °F
12
6-8
110
101
7
90
4
100
12
6-8
11C
95
7
90
4
100
18
6-8
110
70
7
90
2
100
18
6-8
110
90
7
90
2
100
Regeneration Steam
Steam Rate, lb/lb solvent
Steam Pressure, psig
10
15
10
15
10
15
10
15
Cycle
Adsorbing, hour
Desorbing, hour
Cooling, hour
14
6
1
14
6
1
7 5
5 0
2.5
Stream Time
Hour9/day
Days/week
Weeks/year
20
6
46
20
6
46
20
6
46
20
6
46
-58-
-------
RADIAN
CORPORATION
Air Coolers
Air coolers are required in the oven emission control
systems to cool the air from 375°F to 100°F prior to entering the
adsorbers. Each adsorption system contains 2 water-cooled,
floating-head, shell-and-tube air coolers. They are constructed
of carbon steel and each has a heat exchange surface of 1500
square feet.
No air coolers are required to pretreat the air from
the paint spray booths.
SLA Blowers
A large blower is required for each of the paint spray
emission control systems. For the larger system, the blower
is sized to handle 350,000 scfm with a head of 17 inches of water.
The same head is developed by the blower for the smaller system
which is rated at 200,000 scfm.
Each of the smaller oven emission control systems
requires a blower sized to handle 10,000 scfm of air and to
develop a head of 30 inches of water.
All blowers are designed to operate at 80 percent of
rated capacity with an overall efficiency of 65 percent, and all
are constructed of carbon steel.
Incinerator-Boiler
Each of the adsorption systems includes an incinerator-
boiler in which the regeneration stream, containing steam and
desorbed solvent, is burned. The heat produced is used to produce
more steam for regeneration. The units are designed to produce
-59-
-------
RADIAN
CORPORATDON
two to three times the average amount of required steam. The
incinerator-boilers for the spray booth control systems are
rated at 11,000 and 5000 pounds of 15 psig steam per hour. Each
of the units for the oven control systems is rated at 1250 pounds
of steam per hour.
All of the units are designed to use natural gas or
fuel oil as an auxiliary fuel for heatup and control purposes.
Combustion Air Preheater
The combustion air required to incinerate the desorbed
solvent is preheated before mixing it with the steam-solvent
stream to avoid condensing the steam. The source of heat is the
incinerator-boiler flue gas.
The exchangers are relatively small and inexpensive.
Their size ranges from 105 square feet for the largest system
to 15 square feet for the smaller oven emission control system.
They are shell and tube exchangers constructed of carbon steel.
Control Valves
All control valves are of the butterfly type. Forty-
eight-inch valves are required in the spray booth control systems.
Two valves are required on each adsorber, one on the air inlet
and one on the air outlet lines. They must be capable of closing
with minimal leakage against a pressure of 30 inches of water.
Thirty-inch and twenty-four-inch valves are required on
the air inlet and outlet lines, respectively, for the adsorbers
in the oven control systems. They are required to close against
a head of 50 inches of water. All these valves are constructed
of carbon steel.
-60-
-------
RADIAN
CORPORATION
4.4.3 Plot Plan
Figure 4.4-1 is a rough plot plan showing the equipment
arrangement in the solvent emissions control systems for the
typical auto assembly plant. In this configuration, the two
spray paint booth control systems with the vertical adsorbers are
located next to one another. The paint oven control systems,
incorporating horizontal adsorbers, are situated at either end of
the plot.
A total of about 16,000 square feet (0.4 acre) is taken
up by the control system.
-61-
-------
filt
AOSORBE*
6' X 16*
INCIN. -
BOILER
AOSORBER
6' X 18*
ADSORBER
20' OiamETER
AOSORBER
20' oiAucrefi
AOSORBER
29* OIAMETER
BOILER-
INClNERATOR
filt.
BOILER-
INCINERATOR
sua
FAN
A090P8ER
20' OiamCTCR
f A0S0R8ER \
[ 20' OIAMETER 1
AOSORBCR
20' DIAMETER
filt
AOSOR9ER
6* X 18*
INCIN. -
80ILEP
A08QR8£R
6' X 18'
FIGURE 4.4-1
SOLVENT EMISSION CONTROL SYSTEM FOB A TYPICAL AUTO ASSEMBLY PLANT
PLOT PLAN
-62-
-------
RADIAN
CORPORATION
5.0 PROCESS ECONOMICS
Equipment, utility, and operating costs were determined
for emission control systems for the paint booth and oven design
cases defined in Sections 4.2 and 4.3, and for a typical auto
assembly plant as described in Section 4.4. The costs are
enumerated in this section. The sensitivity of the solvent
removal cost to variables such as system size, solvent concentra-
tion, and air velocity was investigated and is also discussed in
this section.
5.1 COST DATA
A number of sources were used in the estimation of
equipment costs. Principal among these were the cost data pre-
sented by Guthrie (GU-075) for various individual equipment items.
Also used were a recent report by McGlamery (MC-136) which gave
detailed equipment cost estimates for S02 removal processes,
Perry (PE-030), Peters and Timmerhaus (PE-065), and estimates
by equipment vendors. The equipment costs were brought up to
date as of July 1975, using the Chemical Engineering Plant Cost
Index.
Utility costs were obtained from several sources
including Perry (PE-030), Peters and Timmerhaus (PE-065) and
Grandjacques (GR-193).
In some cases, equipment and utility costs represented
an average of values obtained from several sources.
5.1.1 Equipment Costs
The purchased costs (f.o.b.) of major equipment items
were obtained from the various sources mentioned above. The
-63-
-------
RADIAN
CORPORATION
installed cost of each major piece of equipment was estimated
using Guthrie's "bare module" factor (GU-075). This factor
represents the average cost incurred when installing specific
types of equipment in plants. Included in the installed cost,
in addition to the purchase price, are all associated piping,
foundation, instrumentation, insulation, field labor, and indirect
costs such as freight, taxes, insurance, construction overhead,
and contract engineering costs.
Adsorber costs were estimated from Guthrie (GU-075) and
included the cost of vessel internals which were priced at one-
half the cost of trays for a distillation column of the same size
as the adsorbers.
The purchased prices (f.o.b.) of the solvent-laden air
blower, incinerator-boiler fan, and drying air fan were based on
blower prices from McGlamery (MC-136). Costs were adjusted for
size differences using Guthrie's cost size exponent of 0.68
(GU-075) •
The costs of the large butterfly control valves were
estimated using information from various literature sources
(GU-068, LI-134) as well as vendor estimates. Cost data for large
(>14") control valves are not readily available, and the values
that were finally used represent an average of cost estimates
from several sources.
Package boiler prices are presented by Guthrie (GU-075).
It was assumed that these costs would suffice for combination vapor
incinerater-boilers after adding 10% to the boiler cost to allow
for minor modifications. Prices were based on steam capacity.
-64-
-------
RADIAN
CORPORATIOH
The combustion air preheaters and air coolers (for
cooling the paint oven air) costs were determined from Guthrie
(GU-075)> anc* were based on heat exchange surface areas.
The costs of air filters were estimated from information
obtained from various sources. Vendor estimates ranged from $10
to $60 for disposable units to treat 2500 cfm. The service time
of these units was assumed to be one month. The initial purchase
and installation costs represent a combination of values obtained
from Perry (PE-030), Guthrie (GU-075), and vendor estimates.
5.1.2 Utility Costs
The utility costs used in this study are listed below:
Natural Gas
Electricity
Boiler Feedwater
Cooling Water
Recoverable heat
$1.00/1000 scf
$.034/kwh
$.50/1000 gal.
$0.10/1000 gal.
$1.00/106 Btu
The cost of natural gas was taken from Grandjacques
(GR-193) and McGlamery, et al. (MC-136). The price of electricity
was also taken from Grandjacques (GR-193), and it is somewhat
higher than that found in other sources. It was used, however,
because it represents information obtained from an operating
solvent recovery plant.
Cooling water and boiler feedwater costs vary widely.
Costs in 1967 were $.02-.08/1000 gal for cooling-tower water and
$0.15-0.40/1000 gal for boiler feedwater (MO-189, PE-065). Grandjacques
(GR-193) used $0.50/1000 gal for process water. In refineries, the
cost of boiler feedwater was reported as $0.15/1000 gal (MO-189).
-65-
-------
RADIAN
CORPORATION
To be conservative, the higher costs for water (adjusted to
1975) were used in developing costs.
Recoverable heat was rated at the same value as that
obtained from natural gas. The value of $1.00/MM Btu is
approximately equivalent to $1.00/1000 scf for natural gas, since
the heating value of natural gasis about 1000 Btu/scf.
5.2 Paint Spray Booth Emission Control Systems
The cost of solvent vapor removal using carbon adsorp-
tion systems is highly dependent on a number of variables. One
of these is the size or capacity of the system. The capital
requirements and operating costs were estimated for spray booth
control systems for the design cases described in Section 4.2 of
this report. The costs were all determined for battery limits
plants installed at ground level in mid-1975. Shown in Table
5.2-1 are the capital and operating costs for Cases I, IIA, IIB,
III (Base Case), and IV. The effects of the system size on
costs are investigated in these cases.
In Cases I and IIA, horizontal adsorbers are used,
while vertical vessels are employed for the larger systems. The
capital costs, including a 30% contingency factor and the initial
cost of the carbon charge, range from $520M for the 60,000 scfm
horizontal system to $4,699M for a system to treat 900,000 scfm
(Case IV) using vertical adsorbers. Three major equipment items
make up most of the capital costs. By far the largest cost items
are the adsorbers, which account for about 50 percent or more
of the total installed equipment cost. The relative costs of
the solvent-laden air blowers and the large control valves vary
according to the size and configuration of the system. The cost
of the SLA blowers ranges from $39.5M for Case I to $302M for
Case IV, and rises with increasing system capacity. The cost of
-66-
-------
TABLE 5 2-1
SOLVENT EMISSION CONTROL BY CARBON ADSORPTION/INCINERATION
SPRAY BOOTH SYSTEMS
CAPITAL
AND OPERATING
COSTS (BATTERY
LIMITS PLANT)
CASE
CASE
I
IIA
IIB
III
(Base Case)
IV
Solvent
Toluene
Toluene
Toluene
Toluene
Toluene
Solvent concentration, ppn
50
50
50
50
50
Air Flow Race, scfm
60,000
250,000
250,000
450,000
900,000
Air Velocity Through Bed, ft/min
110
110
110
110
110
Capital Cost (Installed, 1975), $
Adsorbers
233.700
779,000
1,021,200
1,599,000
2,666.000
SLA Blowers
39,500
102,000
102,000
151,000
302.000
Other Blowers
4.000
7,000
7.500
9,500
9,500
Boiler/Incinerator
12,500
18,000
18,000
33.000
52,000
Control Valves
74,400
248,000
120,000
120,000
240,000
Air Preheater
6,000
11,000
19,000
24,000
24,000
Filters
10.500
25,000
25.000
50,000
90,000
380,600
1,190,000
1,312,500
1,986,500
3,384,000
Contingency (301)
114,200
357,000
393,800
597,000
1,015.000
494,800
1,547,000
1,706.300
2,583,500
4.399,000
Carbon
-------
RADIAN
CORPORATION
the control valves is dependent on the configuration of the
system as well as the size. For example, the cost of the valves
for Case IIA (250,000 scfm) is $248M, while in Case III (450,000
scfm) the value is only $120M. This is due to the use of 10
horizontal adsorbers, requiring 20 valves for Case IIA, as
opposed to 3 large vertical adsorbers and 6 valves for Case III.
The effect of the system size on the capital cost is
presented graphically in Figure 5.2-1. In terms of system
capacity, the capital cost ranges from $5.22 to $8.67/scfm. The
cost decreases as the size of the system increases.
The annual operating costs range from $174M for the 60,000
scfm size system (Case I) to $1,554M for the 900,000 scfm (Case IV).
The operating labor requirements range from 0.1 to 0.5 man/shift
for Cases I and IV, respectively. Maintenance was set at 4% of
the total capital cost and includes monthly replacement of the
filter medium. The capital cost (excluding the cost of the carbon)
is amortized at an annual interest rate of 12 percent over a ten-
year period.
As might be expected, the capital charge represents the
highest single operating cost. The next highest operating cost
is electricity, which is used almost entirely to drive the SLA
blowers. Thus, it can be seen that any increase in required
carbon bed depth and associated pressure drop through the system
will be very costly in terms of operating costs. If the bed
depth is doubled from 12 inches to 24 inches, the cost of elec-
tricity to power the SLA blowers would probably exceed the capital
charges for the larger systems.
The solvent removal costs vary from $0.80 to $0.51 per
pound of solvent removed, with the cost declining as the size of
-68-
-------
2
Ll.
(J
V)
N
4»
CO
o
o
< 6
t-
0.
<
o
SOLVENT: TOLUENE
SOLVENT CONCN: 50 VPPM
HORIZONTAL BEDS
VERTICAL BEDS
I
_L
_L
X
JL
J-
100 200 300 400 500 600 700
AIR FLOW RATE. I03 SCFM
800 900
1000
FIGURE 5.2-1
CARBON ADSORPTION SYSTEM FOR SPRAY BOOTH SOLVENT EMISSION CONTROL
CAPITAL COST: EFFECT OF AIR RATE
-69-
-------
RADIAN
CORPORATION
the system becomes larger. In terms of system capacity, the
removal costs range from $2.89 to $1.85 per year per scfm. The
solvent removal costs as a function of system size are shown in
Figures 5.2-2 and 5.2-3.
5.3 Paint Oven Emission Control Systems
The solvent removal costs were determined as a function
of the system size or capacity for paint oven emissions. The
capital and operating costs are presented in Table 5.3-1 for four
cases (Cases IX-XII) with the size of the systems ranging from
10,000 scfm to 250,000 scfm. All costs were calculated for
battery limits plants installed at ground level in 1975.
In comparing the capital costs associated with oven
emission controls to those estimated for spray booth systems, it
is immediately apparent that oven control systems are much cost-
lier. The capital costs range from $383M to $3,309M for systems
processing solvent-laden air volumes of 10,000 to 250,000 scfm,
respectively. By comparison, a capital cost of $1,531M is required
to treat 250,000 scfm of air from spray booths (Case IIA). There
are several reasons for the higher costs. It requires more equip-
ment to process air from paint ovens. The air must be cooled
and the air coolers are quite large and very expensive. The
solvent concentration in the oven air is 300 vppm as compared to
50 vppm used in developing costs for spray booth control systems.
The higher concentration results in the need for larger boiler-
incinerators, deeper beds, and larger SLA blowers. The capital
cost is shown as a function of the system capacity in Figure 5.3-1.
The operating costs are also much higher for oven control
systems. Part of the greater cost is due, of course, to the higher
capital charge. Another very large operating cost, however, is the
cooling water required to cool the incoming air. This cost is
-70-
-------
0.8C
>
o
2
c 0.70
CO
—I
CO
o
o 0.60
_i
<
>
o
2
Ul
-------
3.00
2.80
2.60
a:
> 2.40
1
2
u.
o
" 2.20
^ 2.00
O
~ 1.80
<
(£
Li
O 1.60
1.40
1.20
1.00
X
T
T
T
SOLVENT: TOLUENE
SOLVENT CONCN: 50 VPPM
HORIZONTAL BEOS
VERTICAL BEDS
100 200 300 400 500 600 700
AIR FLOW RATE. I03 SCFM
_1_
800 900 1000
FIGURE 5.2-3
CARBON ADSORPTION SYSTEM FOR SPRAY BOOTH SOLVENT EMISSION CONTROL
OPERATING COSTS: EFFECT OF AIR FLOW RATE
-72-
-------
TABLE 5 3-1
CARBON ABSORPTION SYSTEMS CAPITAL AND OPERATING COSTS (BATTERY LIMITS PLANT)
PAINT BAKE OVENS
CASE
IX
XI
XII
Solvent
Toluene
Toluene
Toluene
Toluene
Solvent Concentration, vppm
300
300
300
300
Air Flow Race, scfm
10,000
20,000
50.000
250.000
Air Velocity Through Bed, Ft/tnin
110
110
110
110
Capital Cost, $
Adsorbers
70,600
107,000
181,800
855,900
SLA Blower
25,500
40,900
74,800
284,000
Air Coolers
146,600
190,600
317,800
828,000
Boiler/Incinerator
12,000
19.400
14,300
111,900
Control Valves
21,000
45,000
68,000
270,000
Air Preheater
2.900
4,500
9,300
20,100
Filters
6,100
9,200
16,000
41,900
Other Blowers
2,900
4.600
7,100
17.000
287.600
421,200
689,100
2,428,800
Contingency, 30%
86.300
126,400
206.700
728,600
373.900
547,600
895,800
3,157,400
Carbon @ $1.0C/lb
9.000
16.800
31.500
151.200
Total Capital
382,900
564.400
927,300
3,308,600
Operating Cost, $/vear
Utilities Fuel @ $1.00/
1000 sc£
900
1,600
1.900
24,800
Electricity @
$.034/kwh
11.200
22,400
54.700
285,600
Boiler Feedwater
@ $0.50/1000 gal
200
300
800
3,500
Cooling water @
?0.10/1000 gal
13.900
27,800
69,300
348.200
Recoverable heat
@ $1. 00/MH Btu
(1,300)
(2,600)
(4,400
(33,800)
Operating Labor @ $10.00/man
hour
5,500
5,500
11,000
27,500
Maintenance @ 5Z of capital
19,200
28,200
46.400
165.400
Carbon Replacement (2-year life)
4,500
8,400
15.800
75.600
Capital Charge @ 12% Interest
69.200
101.300
165.700
584.100
Total Operating Cost
123,300
192.900
361.200
1.480,900
Solvent Removed, lb/vear
179,000
432,300
1.080.000
5.420,000
Solvent Removal Cost
$/lb
0.69
0 45
0 33
0 27
$/scfm-year
12.33
9 65
7 22
5 92
Capital Cost, $/scfm
38.29
28 22
18 55
13 23
-73-
-------
40
36 "
32
28
24 -
20 -
12 .
40
SOLVENT: TOLUENE
SOLVENTCONCN: 300VPPM
80
120
160
200
240
AIR FLOW RATE, I03 SCFM
FIGURE 5.3*1
CARBON ADSORPTION SYSTEM FOR PAINT OVEN SOLVENT EMISSION CONTROL
CAPITAL COSTS: EFFECT OF AIR RATE
-74-
-------
RADIAN
CORPORATION
directly proportional to the amount of air treated by the system.
The electrical power is used almost exclusively to drive the SLA
blowers, and the cost of electricity, like cooling water, increases
directly with the volume of air treated.
Maintenance costs are higher than those for the spray
booths because, in addition to replacing filter mediums, the air
coolers may have to be cleaned often to remove deposits of high
boiling compounds emitted from the pain ovens. If these mater-
ials are not present in the oven air, maintenance costs would be
lowered significantly, and the purchase of only one air cooler
would be required.
The cost of removing solvent vapor from paint oven air
is presented graphically as a function of the capacity in Figures
5.3-2 and 5.3-3. The cost ranges from $0.69 to $0.27 per pound
of solvent removed for systems processing 10,000 to 250,000 scfm
of air, respectively. In terms of capacity, the cost is $12.33
per year per scfm for the system handling 10,000 scfm. This cost
drops to $5.92 per year per scfm for Case XII (250,000).
5.4 Control System for a Typical Auto Assembly Plant
The costs of installing and operating a solvent control
system for the typical auto assembly plant as defined in Section
4.4 were estimated. The costs, summarized in Table 5.4-1, are for
a battery limits plant located at ground level and completed in
mid-1975. The costs do not include any off-site facilities such
as air collection ducts, cooling tower, power station, etc. The
plant is of sufficient size to reduce the solvent emissions to
3000 pounds per day for each spray booth and by 85% for each oven.
The total capital cost for the control system is $4,504M
or $7.90 per scfm. The costliest capital items are the adsorbers,
which account for over 50 percent of the overall cost. The SLA
-75-
-------
SOLVENTS TOLUENE
SOLVENT CONCN: 300 VPPM
-I L.
40
80 120 160 200
AIR FLOW BATE. I03 SCFM
240 280
FIGURE 5.3-2
CARBON ADSORPTION SYSTEM FOR PAINT OVEN SOLVENT EMISSION CONTROL
SOLVENT REMOVAL COSTS: EFFECT OF AIR FLOW RATE
-------
SOLVENT: TOLUENE
SOLVENT CONCNs 300 VPPM
FIGURE 5.3-3
CARBON ADSORPTION SYSTEM FOR PAINT OVEN SOLVENT EMISSION CONTROL
OPERATING COSTS: EFFECT OF AIR FLOW RATE
-------
TABLE 5 4-1
COST OF SOLVENT EMISSION CONTROL SYSTEM FOR TYPICAL AUTO ASSEMBLY PLANT
(BATTERY LIMITS PLANT)
System Type
Paint Spray
Booths
Paint
Ovens
Air Capacicy, scfm
350.000
200.000
10,000
10.000
Assembly Line
A
B
A
B
Totals
Capital Cost, J
Adsorbers
1,250,000
895,000
70,600
70,600
- 2,286.200
SLA Blowers
130,000
91,000
25,500
25,500
272,000
Other Blowers
7,500
6,400
2,900
2,900
19,700
Boiler-Incinerator
22,800
13.000
12,000
12,000
59,800
Control Valve8
120,000
120,000
21,000
21,000
282,000
Air Coolers
-
-
121,400
121.400
242.800
Air Preheater
23,600
19,800
2,900
2,900
49,200
Filters
22.000
6.100
6.100
64.800
1.584,500
1,167,200
262,400
262,400
3.276.500
Contingency @ 30%
475.400
350.200
^8.700
78.700
983.000
2,059,900
1,517,400
341,100
341,100
4,259.500
Initial Carbon Charge, $
144.000
82.500
9.000
9.000
244.500
2,203,900
1.
,599,900
350,100
350,100
$4,504,000
Operating Cost, $/Year
Utilities- Gas
3,600
2,100
600
800
7,100
Electricity
221,000
125,800
7,100
9.200
363,100
Boiler Feedwater
700
400
100
100
900
Cooling Uater
-
-
8,800
9,200
18.000
Energy Credit
(9,900)
(5,700)
(800)
(1.200)
(17,600)
Operating Labor
22.000
22,000
5,500
5,500
55,000
Maintenance
88.200
63,600
17,500
17,500
186,800
Carbon Replacement
36,000
20,600
4,500
4,500
6S.600
Capital Charge Q 127. interest
381.100
280.700
64.800
64.800
791.400
742,700
509,300
108,200
110,500
$1,470,300
Total Solvent Removed, lb/year
1.159,200
607,400
137,900
179.300
2,083,800
Solvent Removal Cost,
$/lb
0.64
0.84
0.78
0.62
0 71
$/scfm-year
2.12
2.55
10.82
11.05
Capital Cost , $/scfm
6.30
8.00
35 01
35.01
7 90
-78-
-------
RADIAN
CORPORATION
blowers, control valves, and air coolers are all approximately
equal in cost.
As might be expected, the capital charge, amortized
over a ten-year period at 1270 annual interest, is the largest
single component of the operating cost. The utilities require-
ments are listed in Table 5.4-2. Of particular interest is the
cost of electricity, which at $363M per year, makes up 25% of the
total annual operating cost. The cost of natural gas at $7.1M
per year is very small compared to that of the electricity. When
the credit for recovered heat is compared to the cost of the
natural gas, it can be seen that, for this plant, incineration
of the solvent provides more heat than is required for steam
generation.
Maintenance cost is relatively high due to the required
periodic replacement of the filter medium and cleaning of the air
coolers to remove condensed material.
The total solvent removed by the control systems amounts
to 2,084,000 pounds per year. The removal cost is $0.71 per
pound or $1420 per ton of solvent removed. In Table 5.4-3 the
cost effectiveness of this system is compared to other methods
of reducing hydrocarbon emissions. It is more expensive than
balance systems for refueling emissions at service stations, but
less costly than motor vehicle emission control methods.
It should be emphasized that no significant attempt was
made to optimize the equipment or plant design in order to mini-
mize costs. When data from individual auto assembly plant paint
operations become available, such optimization might significantly
reduce some of the costs associated with carbon adsorption/
incineration control systems. For example, the power requirements
-79-
-------
RADIAN
CORPORATION
TABLE 5.4-2
SOLVENT EMISSIONS CONTROL SYSTEM FOR HYPOTHETICAL
AUTO ASSEMBLY PLANT
UTILITIES REQUIREMENTS
Natural Gas
Electricity
Boiler Feedwater
Cooling Water
Energy Credit
7.1 x 106 scf/year
10.7 x 106 kwh/year
2.8 x 106 gal/year
180 x 10s gal/year
17.6 x 109 Btu/year
-80-
-------
RADBAENI
CORPORATION
TABLE 5.4-3
COST EFFECTIVENESS OF METHODS AVAILABLE FOR
REDUCING HYDOCARBON EMISSIONS
$/Ton
Controlled
Degreasing Controls 1
Dry Cleaning Controls 29
Painting Substitutions 116
Jet Engine Modification 177
Balance System for Refueling
Emissions at Service Stations 634
Jet Aircraft Towing 894
Solvent Vapor Removal from Air
by Carbon Adsorption** 1,420
Piston-Aircraft Afterburner 3,610
LDMV* Minor Exhaust Retrofit 3,010
LDMV* Evaporative Controls 3,250
LDMV* Catalytic Converter 3,700
LDMV* State Inspection/Maintenance 5,830
Maximum LDMV* Maintenance 16,300
Modest Bus Improvement 13,300
Modest Mileage Surcharge 13,300
Heavy Mileage Surcharge 18,000
Major Bus Improvements 31,000
* LDMV - Light Duty Motor Vehicles
** Cost for typical auto assembly plant as defined in this study.
Costs are based on battery limits plant only.
Source: GO-091, CA-294
-81-
-------
RADIAN
CORPORATION
of the SLA blowers can be reduced through the use of a shallower
bed and/or lower air velocity. The reduction in power requirements
can only be accomplished, however, at the expense of additional
capital. A carbon bed shallower than 12 inches (if practical)
will require more frequent regeneration and, therefore, larger
regeneration equipment. If the air velocity is reduced, more
bed area and larger adsorbers are needed.
As another example, the cost of cooling oven air might
be reduced by using air-cooled instead of water-cooled exchangers.
The additional cost of air-cooled exchangers followed by water-
cooled trim coolers might be outweighed by the savings in cooling
water costs, particularly if new cooling tower capacity is
necessary. Depending on individual plant requirements, the cost
of off-site equipment such as air collection ducts, cooling tower,
electrical substation, etc., could add substantially to the
solvent removal costs. For example, the installed cost of an 8'
by 8' duct is about $100 per linear foot (GU-079). The cost of
off-site equipment could be as high or higher than the cost of
the battery limits plant.
The amount of air to be treated could be minimized by
evaluating the individual air streams and treating only those
which have higher solvent concentrations.
5.5 Sensitivity Study
A wide-ranging study of the sensitivity of the solvent
removal costs to the many variables is beyond the scope of this
work. It was possible, however, to do a brief investigation of
the effects of air velocity and solvent concentration on the
capital and operating costs. This was done for Cases V through
VIII, which were then compared to the Base Case III. The cost
breakdown for these cases is shown in Table 5.5-1.
-82-
-------
TABLE 5 5-1
PAINT SPRAY BOOTH CONTROL SYSTEM CAPITAL AND OPERATING COSTS (BATTERY LIMITS PLANT)
EFFECT OF AIR VELOCITY AND SOLVENT CONCENTRATION
CASE
III
(Base Case)
V
VI
VII
VIII
Solvenc
Toluene
Toluene
Toluene
Toluene
ToLuene
SoLvenc Concentration, ppm
50
50
50
150
300
Air Flow Race, scfm
450,000
450.000
450,000
450,000
450,000
Air Velocity Through Bed. ft/rain
110
80
50
110
no
Capital Cose (Installed. 1975), $
Adsorbers
1,599,000
2,132.000
3,198,000
1,599,000
1.
599,000
SLA Blowers
151,000
130.000
102,400
151,000
183,000
Other Blower
9,500
13,000
13,000
13,000
13,800
Control Valves
120,000
160,000
240,000
120,000
120,000
Air Preheater
24,000
24,000
24,000
16,000
25,000
Filters
50,000
50,000
50,000
50,000
50,000
1,986,500
2.542,000
3,660,400
2,037.400
2.
094,900
Contingency (30%)
597,000
763,000
1,098,100
611,200
623,500
2,583,500
3,305,000
4,748,500
2,648,600
2.
723.400
Carbon @ $1 00/lb
180.000
237.400
356.800
180.000
258.000
Total Capital,
2.763,500
3,642.400
5,105,300
2,828,600
2,
981.400
Operating Cost, 5/Year
Utilities Fuel @ $1 00/
1000 scf
4,900
4,900
4,900
18,000
36.100
Electricity @
?.034/kvh
283,900
224,400
157,400
283,900
368,200
Boiler Feedwater
@ $0.50/1000 gal
1,000
1,000
1,000
3,200
6,500
Recoverable Heat
@ $1.00/MM Btu
(6,600)
(6,600)
(6,600)
(33,500)
(61,600)
Operating Labor @ $10 00/Man-
Hour
27,500
27,500
27,500
27.500
27,500
Mainteaance @ 47. of Capital
Cost
111,000
146,000
204,000
113.000
119,000
Carbon Replacement (4-Year
Life)
45,000
59,400
89,200
45,000
64,500
Capital Charge @ 127. interest
478.600
611.400
878.500
490.000
503.800
Total Operating Cost
945,300
1,068,000
1,355,900
947.100
1.
064,000
Solvent Removed, lb/year
1.630,000
1,630,000
1,630,000
4,891,000
9,
757 ,000
Solvent Removal Cost,
S/lb
0 58
0 65
0 83
0 19
0 11
$/scfm-year
2 10
2 37
3 01
2 10
2 36
Capital Cost, 5/scfm
6 15
8 09
11 35
6 29
6 62
-83-
-------
RADIAN
CORPORATION
The effect on costs of reducing air velocity through
the carbon bed in a spray booth control system is presented
graphically in Figures 5.5-1 and 5.5-2. Except for the air
velocity, the conditions are the same as those of Case III (Base
Case) which were described in Section 4.2. As the allowable
velocity decreases, the required bed face area increases, thus
increasing the size of the adsorbers, which are the costliest
equipment items in the system. As shown in Figure 5.5-1, the
capital cost rises from $6.15 per scfm at an air velocity of 110
fpm to $11.35 per scfm at an air velocity of 50 fpm. The effect
of the air velocity on the solvent removal cost can be seen in
Figure 5.5-2. As the velocity is decreased from 110 to 50 fpm,
the cost of removing solvent from the air rises from $0.58 to
$0.83 per pound of solvent removed. The increase in operating
cost due to the higher capital charge is partially offset by the
decrease in the power consumption by the SLA blower. For example,
the cost of electricity drops from $284M to $159M per year as
the velocity decreases from 110 to 50 fpm.
The effect of increased solvent concentration on capital
costs is not large, at least in the range of 50 to 300 vppm con-
centration. This is evident from Figure 5.5-3. The capital cost
is $5.15 per scfm at a toluene concentration of 50 vppm and rises
to only $6.62 per scfm at 300 vppm, an increase of less than 8
percent. The cost of the adsorbers is the same for all solvent
concentrations within the range of the study. The increase in
capital requirements at the higher concentrations comes from the
larger SLA blowers required to overcome the increased resistance
across the deeper bed.
The operating cost rises from $945M to $1,064M per year
as the solvent concentration increases from 50 to 300 ppm. More
fuel is needed at the higher concentrations for control purposes
and bed heatup due to the shorter regeneration time. However,
-84-
-------
o
CO
\
co
o
o
0.
<
o
13.00
12.00
11.00
10.00
9.00
8.00
7.00
6.00
5.00
AIR FLOW RATE: 450,000 SCFM
SOLVENT: TOLUENE
SOLVENT CONCNi 50 VPPM
30 40 50 60 70 80 90 100 110 120 130
AIR VELOCITY, FT/MIN
FIGURE 5.5-1
CARBON ADSORPTION SYSTEM FOR SPRAY BOOTH SOLVENT EMISSION CONTROL
CAPITAL COSTS: EFFECT OF AIR VELOCITY
-85-
-------
t-
co
o
o
0.90
CO
_l
\
0.80
AIR FLOW RATE: 450.000 SCFM
SOLVENT: TOLUENE
SOLVENT CONCN: 50 VPPM
<
>
o
2
bl
a.
0.70 -
o
o
0.60 -
0.50
30
40
50
60
70
80
90
100
110
120
AIR VELOCITY THROUGH BED. FT/MIN
FIGURE 5.5-2
CARBON ADSORPTION SYSTEM FOR SPRAY BOOTH SOLVENT EMISSION CONTROL
SOLVENT REMOVAL COST: EFFECT OF AIR VELOCITY
-------
I
00
I I
o
to
\
-------
RADIAN
CORPORATION
all this heat, and more, is recovered. The power requirements
are somewhat higher at the 300 vppm concentration. This is due to
the need for a larger blower as discussed in the preceding para-
graph. As illustrated in Figure 5.5-4, the solvent removal cost
declines from $0.58 to $0.11 per pound as the solvent concentration
increases. Although the operating cost rises somewhat, the
amount of solvent removed increases from 1.630M to 9,757M pounds
per year, resulting in decline in the removal cost on a per-
pound basis.
Certainly, as the solvent concentration is increased
beyond 300 vppm the operating cost will continue to rise. When
the concentration gets high enough, more adsorbers will be
required and at this point the capital cost will rise more
significantly.
The solvent concentration in the regeneration stream from
the adsorbers will vary during the regeneration cycle. If the
variation is too great to permit efficient operation of the incinera-
tor for direct vapor incineration, the solvent/steam mixture can
be condensed. The liquid solvent and water can then be incinerated
at a steady rate. Additional operating costs would be incurred
primarily from the increased use of cooling water and incinerator
fuel. For example, the total operating costs for Case III (Table
5.5-1) would be increased by 10-15% over those for the noncondensing
mode of operation.
-88-
-------
0.65
i
00
vO
m
\
C/J
o
o
_J
o
3
Z
UJ
>
o
-------
RADIAN
CORPORATION
6.0 ENERGY CONSIDERATIONS
The carbon adsorption/incineration system effectively
concentrates the solvent for a more efficient incineration. Some
fuel will be required, however, to start the regeneration and to
provide some control of the steam rate during the regeneration.
The fuel requirements, availability, and the subsequent recovery
of the combustion/incineration heat are discussed in the following
sections.
6.1 Fuel Requirements
The fuel requirements for the adsorption/incineration
system are quite small within the range of process and operating
conditions selected as bases for the various design cases in
this report. The control system for the typical auto assembly
plant, for example, requires only 7.1 x 109 Btu per year in sup-
plemental heat. This is equivalent to about 7.1 x 106 scf of
natural gas or about 50,000 gallons of fuel oil. By comparison,
about 32 x 109 Btu per year are required to heat 20,000 scfm of
air from 75°F up to 375°F as in a bake oven operation.
Actually, the heat produced from burning the regenera-
tion stream of toluene and steam is more than sufficient to
manufacture the steam required in the regeneration. The supple-
mental fuel is necessary to initially heat the adsorber and
carbon bed from the adsorbing temperature of 100°F or less up
to the regeneration temperature of 220°F. In addition, some
supplemental fuel will undoubtedly be necessary in order to
control the regeneration steam rate, since the output of solvent
vapor from the adsorber will not be steady during the regeneration.
-90-
-------
RADIAN
CORPORATION
6.2 Fuel Availability
The availability of fuels is dependent on both the
type of fuel required and the location of the plant. Natural
gas is in short supply for industrial use in many areas. The
supply is usually tightest in the northern and eastern sectors
of the U.S.
If natural gas is to be used as the supplemental fuel
source in carbon adsorption/incineration systems it should be
available to the auto assemply plant under supply conditions
similar to those already in existence at the plant. For example,
if the gas being used at the plant is delivered on an interruptible
basis, the bake ovens will not be operating during periods of
interrupted supply, the paint assembly line will be shut down,
and the solvent removal system will not be needed.
In some areas, however, it may be impossible to obtain
natural gas for new facilities, or it may be available under
conditions which would make its use impractical. In these cases
it would be advisable to consider other fuels such as fuel oil.
As discussed in Section 6.1, about 50,000 gal/year of fuel oil
would be needed to supply the supplemental fuel requirements of
the adsorption systems for the typical auto assembly plant. This
would require a minimal amount of tankage and additional equip-
ment .
6. 3 Energy Recovery
At a steam/solvent ratio of 15-20 pounds of steam per
pound of adsorbed solvent, enough solvent is present (at average
rate of solvent desorption during regeneration) in the regenera-
tion vapor stream to sustain combusion, i.e., to maintain an
incinerator temperature of 1400-1500°F. At these steam/solvent
-91-
-------
RABM&N
CORPORATION
ratios, more heat than needed to produce regeneration steam is
provided by the combustion of the solvent. In the design cases of
this study, additional heat is supplied from the supplemental
fuel used to initially heat the bed and for control purposes.
Even after producing enough steam for regeneration, a considerable
amount of energy is still available in the form of hot flue gas.
The flue gas leaves the incinerator-boiler at a temperature of
about 750-800°F. The flue gas is used to preheat the combustion
air from ambient temperature up to 220°F, but is available at the
preheater outlet at 700-750°F.
Part of the heat from the flue gas can be recovered.
In the economic evaluation of the design cases, it was assumed
that the gas would be cooled to 300°F, and the heat recovered.
The heat recovery can be accomplished in several ways. If the
location of the carbon adsorption system is close enough to the
bake ovens, part or all of the hot flue gas could be used to
heat the incoming air to the bake oven by employing a gas/air
heat exchanger.
Another means of recovering the excess available heat
is simply to produce additional steam in the incinerator-boiler.
The amount of steam in excess of that required in the regenera-
tor could be used in other plant operations such as space heating.
Higher pressure and/or superheated steam could be produced
separately in the incinerator-boiler, and this could be used to
heat incoming air to the bake ovens, particularly if the ovens
are located some distance away from the adsorption system.
One of the major problems in recovering the heat,
however, will undoubtedly be due to the unsteady solvent output
during regeneration of the carbon bed. The solvent concentration
in the regeneration stream initially is low, rises to a peak, and
then continually declines until the regeneration is complete.
-92-
-------
RADIAN
CORPORATION
The heat output from burning the regeneration stream varies with
the solvent concentration. Methods of reducing the variation in
solvent output during the regeneration will be needed if the
maximum amount of heat is to be recovered. If steam is produced
in parallel with other plant boilers, fuel can be backed out of
other boilers to accomodate cyclic operation.
-93-
-------
RADRAEM
CORPORATION
7.0 OTHER CONSIDERATIONS
Some other aspects of the application of carbon adsorp-
tion systems to auto assembly plants are considered and discussed
in this section. Included are land requirements, plant size, and
plant operation.
7.1 Plant Size/Land Requirements
Carbon adsorption systems having the capacity to treat
the large volumes of solvent-laden air emitted from auto paint
plant operations are physically large. As far as could be
determined, regenerable units of over 250,000 scfm capacity have
not yet been built or operated.
The solvent recovery system designed for the typical
auto assembly plant defined in this report requires approximately
0.5 acre within the battery limits of the unit. The height of
the tallest equipment is in the range of 60-70 feet. Finding an
area of this size convenient to the painting operations will be
a significant problem in many plants where retrofit of carbon
adsorption systems is considered.
In some areas it might be more practical to separate
the treatment facilities in order to locate them more conveniently.
The size of the spray booth treatment systems in almost all cases
precludes their installation on or above the roof of the assembly
plant close to the solvent-laden air outlet. It is conceivable
that the smaller oven control systems could be located on the
roof but a considerable amount of structural modification would
have to be made to provide adequate support.
-94-
-------
RADIAN
CORPORATION
7.2 Plant Operations
The operation of a large carbon adsorption system would
be a familiar and relatively simple task in a chemical or oil
refining plant. In general, however, this is not the case in
auto assembly plants. The personnel are relatively unfamiliar
with the type of large-scale operations involved, such as adsorp-
tion and steam regeneration. It would probably be necessary to
assign men solely to operating and maintaining the systems. It
was assumed that one full-time operator would be required at all
times to operate the system designed for the typical auto assembly
plant. In addition, maintenance crews familiar with the plant
operation and equipment would be required for both routine and
non-routine plant maintenance. Routine maintenance would include
replacing air filters and cleaning heat exchangers.
-95-
-------
RADIAN
CORPORATION
8.0 SUMMARY
The conclusions derived from this study are given in
Section 2.0 of the report. Briefly, it appears that the applica-
tion of large-scale carbon adsorption systems to remove solvents
from auto assembly plant painting operations is technically
feasible. The cost of this method is estimated to be above that
of balance systems for controlling refueling emissions and below
those of methods for controlling hydrocarbon emissions from motor
vehicles.
The required systems are large. In many cases, they
would be larger than any such systems now in existence. In order
to obtain detailed and definitive design data, pilot studies would
be necessary at the individual plant sites.
-96-
-------
REFERENCES
Cannon, T., "Energy Recovery from Solvent Vapor,"
Pollution Engineering, November 1974.
Cavanaugh, E. C., Clancy, G. M., and Dzierlenga, P. S.,
Final Report. Comparison of Stage II Vapor Control Systems,
Report 200-124, Radian Corporation, Austin, Texas,
December 19, 1975.
Chass, R. L., Kanter, C. V., and Elliott, J. H.,
"Contribution of Solvents to Air Pollution and Methods
for Controlling Their Emissions," J. of the Air
Pollution Control Assoc., 13(2), 64-72(1963).
Danielson, J. A., ed., Air Pollution Engineering Manual,
2nd ed., AP-40, Research Triangle Park, N.C., EPA, Office
of Air and Water Programs, 1973.
England, H. M., and Beery, W. T., A Critical Review of
Regulations for the Control of Hydrocarbon Emissions
aria Odorous Pollutants, Air Pollution Control Association,
Pittsburg, PA, June 1975.
Goeller, Bruce F., et al., San Diego Clean Air Project,
Summary Report, R-1362-SD-(Appendix I), Santa Monica,
CA, Rand, 1973.
Grandjacques, B., "Air Pollution Control and Energy
Savings with Carbon Adsorption Systems," Calgon
Corporation Report No. APC 12-A, July 19, 1975.
Guthrie, K. M., "Pump and Valve Costs, " Chemical Eng. 78
(23), 151-159 (1971).
Guthrie, K. M., Process Plant Estimating Evaluation and
Control, Craftsman Book Co., Solana Beach, CA, 1974.
Lee, D. R., "Adsorption Systems and Their Application to
Air Pollution Control," chapter from Control of Gaseous
Emissions, Air Pollution Control Manual of EPA, Durham,
ITT
Lee, D. R., "How to Design Charcoal Adsorption Systems
for Solvent Vapor Recovery," Heating, Piping & Air
Conditioning, 80-83 (May 19707"!
Lee, D. R., "Activated Charcoal in Air Pollution Control,"
Heating, Piping & Air Conditioning, 76-79 (April 1970).
-97-
-------
RADIAN
CORPORATION
LI-134 Liptak, B. G., "Safety Instruments and Control Valve
Costs," Chemical Engineering 77, 94-100 (Nov. 2, 1970).
MA-050 Mattia, M. M., "Process for Solvent Pollution Control,"
Chemical Engineering Progress 66(12) , 74-79 (1970).
MA-491 Manzone, R. R., and D. W. Oakes, "Profitably Recycling
Solvents from Process Systems," Pollution Engineering 5
(10), 23-24 (1973).
MC-136 McGlamery, G. G., et al., Detailed Cost Estimates for
Advanced Effluent Desulfurization Processes, EPA-600/
2-75-006, Tennessee Valley Authority, Muscle Shoals,
AL, January 1975.
MO-189 Monroe, L. R., "Process Plant Utilities," Chemical
Engineering 77, 130-146 (Dec. 14, 1970).
MS-001 MSA Research Corporation, Hydrocarbon Pollutant Systems
Study, MSAR 72-233, AC-860, Evans City, PA, October 1972.
MS-004 MSA Research Corporation, Package Sorption Device System
Study, EPA-R2-73-202, Evans City, PA, April 1973.
PE-030 Perry, John H., ed., Chemical Engineers' Handbook,
McGraw-Hill Book Company, New York 1963.
PE-065 Peters, M. S., and Timmerhaus, K. D., Plant Design and
Economics for Chemical Engineers, McGraw-Hill Book
Company, New York 1968.
RO-227 Roberts, R. E., and J. B. Roberts, "An Engineering
Approach to Emission Reduction in Automotive Spray
Painting," Presented at the 67th Annual Meeting of the
Air Pollution Control Association, Denver, CO, June 9-13,
1974.
VU-003 Vulcan Cincinnati, Inc., "Solvent Recovery Installations,"
Sales and Technical Information Brochure.
WO-070 Wohlers, H. C., and Feldstein, M., "Investigation to
Determine the Possible Need for a Regulation on Organic
Compound Emissions from Stationary Sources in the San
Francisco Bay Area," J. of the Air Pollution Control Assoc
15(5), 226-229 (1965):
-98-
-------
RADIAN
CORPORATION
APPENDIX A
COMMENTS BY PARTICIPANTS IN WORKSHOP
HELD ON 26 FEBRUARY 1976 AT THE EPA
REGION IX OFFICES IN SAN FRANCISCO, CALIFORNIA
-99-
-------
CORPORATION
EPA, Region IX - Workshop held on 26 February 19 76
List of Attendees
Name
Organization
Ron Mueller
Fred Thoits
Karen E. Kohnerts
Jim Loop
Peter Venturini
Bob Roberts
Bob Atkinson
Rod.Buttacavoli
Mike Tyro
Bill Meyer
Don Lee
Bill Thurston
Martin DeVries
Del Twilley
Vic Sussman
Dick Hanselman
Jim Kellt
Allen Koeppel
Bill Krenz
Gary Kendall
Bob Wetherold
EPA Region IX, Air L Haz Mat Div.
If H ii
it II
EPA Region IX, Enforcement Div.
California Air Resources Board
(I M II II
DuPont, Wilmington, Del.
DuPont, Burlington, Cal.
GMC, Assembly Division
GMC, Environmental Activities
Vulcan Cincinnati, Inc.
Vic Manufacturing Co.
EPA Region IX, Enforcement Div
Ford Motor, San Jose, Cal.
" " " Dearborn, Mich.
Union Carbide
So. California APCD
Bay Area APCD
Radian
-------
RADIAN
CORPORATION
COMMENTS BY WILLIAM R. MEYER,
VULCAN CINCINNATI, INC.
-------
'ThO.
2SOO VERNON PLACE
CINCINNATI, OHIO
43219
513 261-2800
TLX 31-^303
March 10, 1976
Radian Corporation
8500 Shoal Creek Boulevard
P. 0. Box 9948
Austin, Texas 78766
Attention: Mr. R. G. Wetherold
Subject: Draft Final Report
Gentlemen:
We thank you and the U.S. EPA for the opportunity to participate
in the February 26th San Francisco workshop to discuss the subject
report. Herein and attached are our comments on the report. We
understand that these remarks may be included in the final report
or appended to it.
Our general impression is that activated carbon solvent recovery
is technically feasible for the auto assembly plant hydrocarbon
emissions detailed in the study as long as present EPA solvent
regulations, especially the 3,000 lb/day maximum discharge to at-
mosphere limits, are not tightened up.
Paint spray booth emission control will be difficult to maintain
at design efficiency levels for long term operation due to the
likely accumulation of undesirable material on the activated car-
bon and the reduced capacity of the carbon when low solvent con-
centrations are in the incoming stream.
We question the assumption that the combined desorption steam and
solvent vapor can be used directly and continuously as fuel for a
waste boiler. About 70% of the system's energy requirement is for
the blower which could be driven by a turbine (continuous demand).
About one-half of the steam used for desorption will condense
during heat-up of the vessel and the carbon, and exit the adsorber
through its steam trap. Further, the regeneration of solvent is
intermittent and lags the demand for regeneration steam by 20 to
30 minutes. This problem can be resolved but the Radian study
does not show the cost of doing it, although it does recognize the
need for additional planning in this area. The anticipated Btu
EPA Contract No. 68-02-1319
Task No. 46
CHEMICAL PROCESS TECHNOLOGY
NCW YORK
CHICAOO
DENVER
LOS ANGELES
-------
Radian Corporation
March 10, 1976
Page Two
Vulcan
value of the recovered solvent will be sufficient to desorb the
system and drive the blower, i.e., the system could be nearly
self-sufficient for its energy requirements. Capital cost would
be increased because of the sophistication of this equipment.
One of the conclusions of this study is that pilot studies be made,
however, it was brought out at the workshop that certain studies
have already been made.
The subject divides itself into two basic cases: 1) paint spray
operations, and 2) baking ovens. There seems little reason not to
proceed on a full scale, say 50,000 CFM, demonstration of solvent
recovery from bake ovens while the paint spray operation proceeds
in pilot stage. This could be worked as a joint industry effort.
EPA may be able to provide funds to assist in engineering and
equipment procurement for the initial demonstrations and evalua-
tion of the systems. The resulting information would be useful to
all involved.
Vulcan's services can be made available in several ways, i.e., for
design, for design and supply, and for design, supply and to super-
vise installation. The total work can be divided into discrete
phases with "go-no go" decisions being made between the phases
which are identified as follows:
Phase I - Vulcan develops a preliminary design for several alterna-
tives with budgetary estimates.
Phase II - After the preferred alternative is chosen, Vulcan de-
velops a definitive process design, an appropriation type
estimate and a detailed scope-of-work.
Phase III - If the cost indicated in Phase II is accepted, and Vulcan
is authorized to proceed, the detailed process and mechani-
cal designs and equipment specifications are prepared.
Phase IV - With authorization to proceed with procurement, Vulcan
provides the purchasing, expediting and inspection ser-
vices, preparation of sub-contract bid packages and evalu-
ation of sub-contractors' bids. Orders nay be placed
directly by Vulcan or for the owner's account at cost,
depending on his preference.
Phase V - Vulcan provides competent erection and advisory field
services relating to mechanical testing, commissioning
and initial operation.
-------
Radian Corporation
March 10, 1976
Page Three
Phase VI - Vulcan, in cooperation with the industry, devises log
sheets, gathers data and evaluates the performance of
the systems.
In this manner, the owner is under no obligation to proceed beyond
any phase or, if he wishes, may authorize more than one phase at a
time.
The + 30% cost estimates of the Radian study need refining, as do
the process schemes and utility estimates. This suggested program
will provide accurate costs, capital and operating, and the process
will be demonstrated.
These are our comments and suggestions. We are interested in learn-
ing the opinions of those who receive this.
WRM:red
Attachment
cc: Mr. Ron Mueller
United States Environmental Protection Agency
Region IX
100 California Street
San Francisco, California 94111
Mr. M. J. Tyro, Staff Engineer
Plant and Environmental Engineering
General Motors Corporation
General Motors Technical Center
Warren, Michigan 48090
Mr. Victor Sussman, Director
Stationary Source Environmental Control
Ford Motor Company
One Parklane Boulevard
Dearborn, Michigan 48126
Very truly yours,
VULCAN CINCINNATI, INC.
William R. Meyer
Manager of Sales,
Environmental Projects
-------
March 10, 1976
Following are some impressions of the Radian study:
Page Par. Process Design
6 7 We would expect the steam/solvent ratio to be less than
5:1 and probably 3:1, not 15:1.
20 2 We fail to understand how a water wash system will not
remove water miscible solvents from a solvent laden air
stream containing water miscible solvents.
27 2 The use of superheated steam to desorb high boilers may
not be desirable. Heat transfer from superheated steam,
or hot gas, is not good when compared to desuperheated
steam which is not practical to use. Carbon will oxi-
dize rapidly in a hot environment. The valves usually
provided for solvent recovery systems desorbing with
desuperheated steam will not withstand the service if
superheated steam is used. More dollars will be added
to the cost of the system for valves. Regenerating the
carbon off site may be the preferred method.
28 3 Mixed solvents do not necessarily reduce the adsorptive
(working) capacity of carbon. Example - Carbon has a
working capacity of about 1% for ethanol and about 7.5%
for a 50/50 mixture of toluene and hexane.
28 4 This statement refers to only a part of the carbon bed,
not the entire bed, i.e., one mass transfer zone. This
paragraph could use clarification.
30 4 The conclusion that the deeper a carbon bed the greater
the is correct but not too significant in vertical,
annular carbon beds which have an inherent low >P. In
vertical or horizontal systems the pressure drop through
the carbon bed is usually 1/3 to 1/2 the pressure drop
of the total system. Deeper carbon beds and the result-
ing less frequent desorption of the adsorber result in
lower steam usage because the sensible heat requirements
of the system are satisfied less often.
32 1 Humidity can be too high or too low. Most solvents will
be preferentially adsorbed in place of water. The mois-
ture remaining on the carbon bed after steaming serves
as a heat sink for the heat of adsorption and this is
desirable. Too low humidity dries the carbon bed pre-
maturely and can cause an increase in the oxidation
(aging) of the carbon and, in extreme cases where ke-
tones are present, fire may result. Adsorption capacity
is lower for hot dry carbon.
-------
Vulcan
Li^/^rvnaA
Page 2 March 10, 1976
Page Par. Process Design
34 2 More steam is used (wasted) during the heat up of the
vessel and carbon than at the end of the steaming
cycle which can be terminated after the "easy" solvent
is removed from the bed. Steam is wasted if an effort
is made to remove too much of the solvent heel. Cer-
tain solvents, such as EtOH, require more desorption
steam than others, such as CCI4. Thick carbon beds,
good volumetric use of the adsorber vessel, as in the
annular vertical design, and high incoming solvent con-
centrations, result in the lowest steam/solvent ratios.
Sensing breakthrough with an analyzer, then initiating
desorption of that adsorber, has proven to be a very
efficient way to economize on steam, however, analyzer
sensing of solvent mixtures which do not remain con-
stant is not practical.
Mechanical Design
25 1 Maximum horizontal adsorber size we have provided is
12' 0 x 50' T/T.
25 2 236,000 CFM can be processed in 5 adsorbers or less.
41 Fresh air fan is not practical in a cold climate. It
can cause freeze up of a wet carbon bed. Better to
take a slip stream from adsorber exhaust. Also, if
this blower is required, the flow should be in the
same direction as steam, causing vapors to flow to
condenser and the condenser should be vented to main
blower inlet.
42 3 The adsorber vessel can be fabricated of carbon steel
by directing the air flow from inside to outside and
steaming outside to inside. If stainless steel is
used for internals, chlorides must be kept to a mini-
mum.
42 4 Blower should be of spark resistant construction with
TEFC drive. Horsepower should be in the range of 30%,
or 50% less than the 2,500 HP drive indicated.
Valves specifid are causing too muchZ\P.
46 3
The incinerator blower will be in a corrosive atmos-
phere and carbon steel is not indicated. More $.
-------
VcTLC^TV .
March 10, 197 6
Mechanical Design
Shell and tube heat exchangers or fin and tube type?
There are ways to design using carbon steel vessels,
either horizontal or vertical type.
Layout
This layout is too tight. Drawings have been provided
to Radian.
-------
RADIAN
CORPORATION
COMMENTS BY W. R. JOHNSON,
GENERAL MOTORS CORPORATION
-------
Environmental Activities Staff
General Motors Corporation
General Motors Technical Center
Warren. Michigan 48090
March 12, 1976
Radian Corporation
8500 Shoal Creek Boulevard
P. O. Box 9948
Austin, Texas 78766
Attention: Mr. R. G. Wetherold
Re: Radian Project 200-045-46 - "Evaluation of a
Carbon Adsorption/Incineration Control System
for Auto Assembly Plants"
Gentlemen:
As requested at the EPA, Region IX, sponsored Carbon Adsorption
Workshop held in San Francisco, California, on February 26, 1976,
General Motors is submitting comments relative to the subject Radian
project report.
General Motors agrees with the concept that a carbon adsorption system
has potential application in controlling solvent emissions from automo-
tive assembly plant paint spray booths and bake ovens. We also agree
on the importance of pilot studies to distinguish in-use feasibility from
theoretical feasibility.
Our comments are not intended to discredit the Radian report but to
bring into perspective the potential complexities of installing and
operating a carbon adsorption system at an automotive assembly plant.
The views expressed herein were either overlooked or treated lightly
in the report and should be considered before any final decision is
made as to the technical and economic feasibility of a carbon adsorp-
tion system as it relates to controlling paint solvent emissions from
automotive spray booths and ovens.
It is acknowledged that activated carbon can be efficiently utilized to
adsorb solvents and be desorbed or regenerated effectively with steam
or by other methods. Treating the total exhausted air from spray booths
and ovens can reduce hydrocarbon (solvent) emissions by approximately
90-95 percent if the carbon adsorption system is properly engineered
initially and regularly maintained.
-------
Mr. R. G. Wetherold
-2-
March 12, 1976
Various automotive paint operations have total exhaust air, containing
hydrocarbons, that range from one million to two million cubic feet
of air per minute. Since carbon adsorption systems are designed
according to the air volume to be handled, this can result in enormous
sizes of equipment with space requirements that most plants will be
unable to provide.
Radian's report circumvents this problem of massive air handling
by stating, or mostly inferring, that only a portion of exhausted air
from each source could be treated to attain the maximum 3, 000 pound
limit imposed by Rule 66. This scheme would be impractical since
it is based on two false assumptions. One is that the 3, 000 pound
limit is permanent. As stated at the workshop, much lower limits
are being considered. Secondly, that hydrocarbon concentrations
remain constant from selected exhaust stacks. This is rarely the
case because of spray booth design and the nature of painting
operations.
There are other factors, either absent or briefly stated in the report,
which must be resolved. Some of these are:
1. Safety requirements - OSHA, NFPA, FM1, etc. have codes for
the protection of workers against fires and potential explosion
hazards. These, or similar codes, may restrict the use of
carbon adsorption systems for solvent collection as designed
for assembly plant paint operation.
2. Instrumentation and controls - reliability and availability require-
ments must be determined.
3. Backup equipment - required in case of failures or breakdowns.
Note: Recently proposed Rule 430 by the Southern California
Air Pollution Control District requires notification to the District
within 30 minutes of an emission control equipment breakdown;
operation of equipment only until end of run or 24 hours, which-
ever occurs first, then it shall be shutdown; if beyond this limita-
tion, a petition for an emergency variance must be filed. Without
such equipment, significant delays and possible indefinite shut-
downs of manufacturing facilities will occur.
4. Natural resources - carbon adsorption systems increase demands
on the availability of electrical energy, natural gas, and oil.
-------
Mr. R. G. Wetherold
-3-
March 12, 1976
5, Battery limits - the ideal conditions upon which the design criteria
and cost figures are based are virtually non-existent at assembly
In addition, there are other operational aspects of a carbon adsorption
system that must be considered and resolved. They are:
a. Pre-filtration of air streams
b. Temperature and humidity controls
c. Fouling of condensers and heat exchangers
d. Noise problems with fans, blowers, etc.
e. Engineering design of valving, locations, booth and oven modifi-
cations, etc.
We believe that the factors enumerated demonstrate the complexities
and unresolved problems associated with the feasible technology of
a carbon adsorption system at an automotive assembly plant. To
avoid inefficiencies, duplications, and costly mistakes, further research
and development work, including practical experience, are required.
plants.
Very truly yours,
W. Ri/Johnson,
W. Ri/Johnson, Director
Plant & Environmental Engineering
MJTrpmm
-------
RADIAN
CORPORATION
COMMENTS BY V. H. SUSSMAN,
FORD MOTOR COMPANY
-------
V. H. Sussman, Director
Stationary Source Environmental Control
Environmental and Safety Engineering
Ford Motor Company
One Parklane Boulevard
Dearborn, Michigan 43126
Mr. R. G. Wetherald
Radian Corporation
March 15, 1976
8500 Shoal Creek Blvd.
P.O. Box 9948
Austin, Texas 78766
Dear Mr. Wetherald:
In accordance with your request at the February 26, 1976 meeting,
the following is an outline of our comments in the order (according
to my notes) in which these points were discussed.
- In this instance, the use of battery limits data in developing
and presenting cost information is inappropriate and misleading.
Everyone at the meeting agreed that these systems would have to
be tailormade to meet specific physical and process differences
at different plants. Such modifications are much more significant
in cost evaluations than battery limits considerations. We
pointed out that, e.g. for a 450,000 CFM "Base Case" system,
we believe that a bare bones cost (i.e. not including humidity,
safety and fire control) would be $9.5 million -- compared.to
the report's $2,763,500 estimate for a battery limits system.
This higher price does not assure that further development
will not be required on parts of the system or that equipment
suppliers would adequately guarantee the removal efficiency
of a system as large as 450,000 CFM. Based upon actual
engineering designs we have performed, we believe that the
difference between battery limits and actual installed system
costs can vary greatly depending upon such factors as:
o incoming electric service, breakers, transformers, water
lines, sewers, cooling tower systems;
o site preparation;
o the cost of duct work from the individual booths and ovens
to the filters. Recent design experience we have had
indicates that a careful analysis must be made of whether
the several exhaust from any booth or oven can be inter-
mingled ahead of the final exhaust to the atmosphere;
-------
Mr. R. G. Wetherald
- 2 -
March 15, 1976
0 the cost of interferences, construction delays, overtime for
final connections and engineering.
1 indicated during the meeting that it appears that the unit
prices for fans, boiler/incinerators, control valves and
filters appear understated for the quality of hardware normally
installed at Ford plants.
- The report did not indicate the criteria used in defining "technical
feasibility." Almost anything is technically feasible, especially
(as I pointed out) to the person who doesn't have to do it. Also,
technical feasibility could be defined as what a handbook on
physical chemistry reports on the potential of various hydrocarbons
with respect to adsorption on and desorption from carbon. Without
a clear definition of "technical feasibility," the report's
conclusions cannot be objective.
- Every place in the report where it is indicated that the proposed
system is technically feasible and/or economically reasonable, a
statement should be made to the effect that this conclusion is
based upon the existence of a Rule 66 "3,000-pound-per-day limit."
As indicated by Bill Krenz, it is almost certain that the report's
conclusions would be significantly different (and, in fact, a
completely new evaluation would be warranted) if the 3,000-pound
limitation were not a prime assumption. I stated that there was
some question as to the regulatory acceptability of the report's
assumptions concerning the manifolding of some of the units to be
controlled. It is questionable as to whether or not a regulatory
agency would permit such manifolding since compliance with emission
limitations could not be determined.
- There was considerable discussion concerning particulate removal
requirements. We showed slides of our Louisville pilot filtration
unit and outlined some of the factors which makes this a major
consideration. The removal of paint particulate matter by filtration
prior to the adsorption stage will be much more costly and
complicated than suggested in the report.
Ford has prepared some concept drawings of carbon adsorption systems
for spray booth emission controls at a specific plant location. One
basic requirement in the Ford design placed the particulate filters
as close to the sources as practical. This approach is taken to
insure a "clean" system of ductwork leading to the adsorber and to
minimize costly cleaning requirements, reduce safety hazards and
structural liability for the large size ducts involved. Radian
concluded that paint particles are "dry" and can be collected in
"disposable fiberglas bag filters" with bag replacement on a
monthly basis. Without additional prefiltration, such a system
would be impractical.
With respect to oven emissions, particulate removal prior to adsorption
and problems resulting from particulate formation on the carbon bed
-------
Mr. R. G. Wetherald
- 3 -
March 15, 1976
were discussed- We believe that the Radian report does not
sufficiently address this issue as well as the issue of the
extensive refrigeration equipment required to cool exhaust gas
prior to adsorption.
Ron Mueller asked if we are planning to provide any filtering
equipment for the new catalytic oven incinerator installation at
Los Angeles. Filtration is not required, because there is no
particulate matter. Emissions are kept in the vapor state and
pass through the burner where they are raised to a temperature
of 8000 f. at which point oxidation can take place in the
platinum impregnated bed.
- Along with our discussions of particulate removal, other issues
related to "exhaust gas conditioning" were also noted.
Except for the brief discussion on page 32, the report did not give
adequate consideration to problems of humidity. Paint spraybooth
exhaust passes through a waterwashed back section in order to remove
entrained particulate matter. Measurements made at typical Ford
plants and observations made on roofs of assembly plants indicate
that the exhaust gases exit the plant at or near saturation. On
particularly cold days, when cooling of the gases takes place in
the stack itself, entrained moisture is observed. In view of the
possible deterioration of the carbon adsorption operation, some
consideration must be included in the design to maintain humidity
control below 80%. During periods of cool ambient temperatures
(less than 700 F.) exhaust gas cooling accompanied by saturation
of the exhaust is highly possible particularly when extremely long
duct runs are used to carry the exhaust gas to a battery limit
adsorption plant. Cost for humidity control could be significant
where extremely large air flows are involved.
- During Bill Meyers' discussion of large adsorption installations, we
raised questions on certain design parameters.
o Although Bill indicated that such installations are possible, we
believe that it is important to recognize some of the significant
design problems involved. For example, the applicability of a
tall vertical annular bed as proposed for the spraybooth systems
is questionable. A 45-foot tall annular column of carbon could
produce considerable forces at the base of the column. This
could cause carbon breakup or attrition resulting in increased
packing density, loss of carbon material, and high pressure
drops. The variation in packing density due to the extreme
bed heights would likely produce nonuniform!'ty in flow through
the carbon bed due to the various degrees of bed void space.
This could result in significant loss of adsorptive capacity
and possibly increased pressure drop throughout the system.
It appears that consideration of carbon adsorption con-
figurations was limited and arrangements were not optimized.
(I would like to re-emphasize that the importance of insuring
uniform air flow through the spraybooth, in to protect
-------
R. G. Wetherald
- 4 -
March 15, 1976
workers and meet OSHA requirements, was not recognized in
some of the proposed systems. The balance of air flow in a
spraybooth system could be greatly affected by discharging
some portions of the booth to the atmosphere and others to a
pressurized system.)
o The use of a design gas velocity of 110 feet/minute at the bed
inlet face appears high. This design velocity would in actuality
produce a face velocity leaving the bed of 126 feet/minute and
could be sufficient to produce carbon attrition, severe bed
packing and high pressure drop. Lower design gas velocities
on the order of 80 feet/minute would be more realistic. This,
of course, would result in significantly larger carbon beds to
accommodate the exhaust air flow.
o The use of 300 ppm as a typical oven concentration is excessive.
For Ford applications 50 ppm as toluene would be more realistic.
While discussing solvent concentration, it should be noted that
care must be exercised in using data from published sources
concerning hydrocarbon emissions and/or control equipment
efficiencies to insure that the basis for measurement is stated.
In most cases the measurement of emissions and/or the control
equipment performance is based on some type of hydrocarbon
analysis using instruments calibrated against standard gases.
As such, the concentrations measured are in actuality related
to some standard gas such as methane or propane. A measurement
of 300 ppm as methane would be equivalent to 50 ppm as hexane
or 43 ppm as toluene. In view of this, it is important when
citing hydrocarbon concentrations to indicate the basis for the
values.
During our presentation on the Ford pilot carbon adsorption study
(an outline description of the study, test results, observations
and conclusions was sent to you with my letter of February 18,
1976), the following points were discussed:
o Oven Tests -- We stated that using the shallow 4" bed in our
test equipment did not achieve the desired 852 reduction. We
concur that a deeper bed would meet this requirement; however,
the necessity for cooling the stack discharge allows condensation
to occur and contaminants will drop out and deposit or cooling
tubes or on the filter media, which creates horrendous disposal
problems.
By operating in the incineration mode, we are back to what we
originally tried to avoid -- the use of an incinerator. The
difference in fuel consumption between a full size oven
incinerator and a somewhat smaller and intermittently operated
carbon adsorption incinerator does not offset the other
installation and operation problems in utilizing a carbon
adsorption system for ovens.
-------
. R. G. Wetherald
- 5 -
March 15, 1976
Operating in the condensate mode creates an even more difficult
problem in disposing of objectionable solvent mixtures and the
contaminated steam condensate. Again, these problems of
disposal and potential water pollution more than offset the
advantage of using this system.
Don Lee of Vic raised the question as to whether our test carbon
bed had not deteriorated as a result of the booth test and there-
fore did not give us optimum results in the oven test. I
believe our Comments and Observations answer this question.
As indicated, the bed was completely regenerated by the use of
super-heated steam; even to the extent of using a minus 20" WC
vacuum pump to insure that the carbon was as good as new.
The question of bed life was also discussed. This probably
would be one of the more significant issues to be resolved by
the proposed pilot test. An important factor in considering
bed life is that high boiling organic vapors cannot be completely
removed from the oven air simply by cooling to 100° F. Some of
these materials are transferred to carbon as aerosol particles.
The high boiling materials cannot be desorbed by use of steam.
Even though the concentration of high boiling materials in the
air stream is small, the total cumulative amount is sufficient
to deactivate carbon rapidly.
(Again, I think it is important to point out that -- (1) We
have had more experience than anyone else in actually testing
these units on automotive painting operations, and (2) If
we had believed that carbon adsorption had potential as a
practical means of meeting Rule 66 oven requirements, we would
not be continuing to use afterburners. As you will recall, we
indicated at the meeting that we are presently installing a
catalytic afterburner as a replacement trial on one of our ovens
in Los Angeles.)
Aside from the obviously fallacious approach of basing costs on
battery limit plants designed to limit emissions to 3,000 pounds
per day, other questions were raised with respect to the "Process
Economics" section.
o Credit is taken in the operating cost of the system, for available
heat in the recovered solvents. This did not appear to be
feasible in the Ford study. The collected solvent vapors are
desorbed in an "avalanche" and(cannot be stored so that)the
heat from the solvents is not available at the time it is
needed in the carbon regeneration cycle.
o Reducing costs to cost per unit is fallacious and meaningless.
The report's operating cost of $1.5 million per year is
projected to a cost of $3.50 per car. For our San Jose Assembly
-------
Mr. R. G. Wetherald
- 6 -
March 15, 1976
Plant, this unit cost is over $9 per vehicle at present
production schedules. This is another example of unqualified,
meaningless statements in the report.
o There are various estimates similar to the table on page 81.
The "$/Ton Controlled" depends upon whether you are starting
from an uncontrolled source or increasing the efficiency of
existing systems. There is considerable disagreement regarding
the control costs listed for a number of the sources in the
table — and we take particular exception to the Radian
report's estimate.
- Some minor p.oints:
0 On page 11 the solvent emissions from spraybooths and ovens
indicates a ratio of approximately 3:1 respectively. Our
measurements on typical Ford enamel and prime spraybooths and
ovens indicates a relationship approaching 9:1. In the L.A.
and Bay areas, I assume many ovens already have emission
control devices in order to comply with existing regulations.
Taking this into consideration the current expected emission
ratio would be well in excess of 9:1.
o On page 13, Table 3.2-1 contains a typical breakdown of solvent
types of various types of coatings. The table does not appear
to reflect typically exempt enamels currently in use in many
areas of the country; too much aromatic compound is present.
On the basis of typical Ford exempt paints, a more likely
breakdown of solvent content would be:
o First sentence, second paragraph, page 20 -- face velocities
should be stated as ft3/min per square foot.
In conclusion, I would like to make some general comments about this
report:
- The report contains a number of gratuitous and cursory comments on
the need for solvent control and the applicability of Rule 66.
(The work statement for this project did not call for addressing
these issues.) If these issues were believed important enough
to be included, then a more complete and detailed discussion
should have been presented. For example:
o In the outline of the provisions of Rule 66; why wasn't the
basis and meaning of the emission standards discussed? The
Aliphatics
Aromatics
Alcohols
Ketones
Esters and Ethers
Nonvolatiles
21%
11%
10%
6%
5%
47%
-------
Mr. R. G. Wetherald
- 7 -
March 15, 1976
3,000 lb/day limitation is a key factor in the report's
evaluation of the required degree (and thus cost) of control.
o In the section on the need to control organic emissions; why
wasn't hydrocarbon reactivity, seasonal variability or inter-
mittent control (if I am allowed to use this term) discussed?
Why wasn't the adequacy of the Bay Area Air Pollution Control
District's regulations (as opposed to Rule 66) in meeting the
"TCP hydrocarbon roll-back task" discussed?
It is the presence of these superficial "throw away" statement
-- not only in the introduction, but also (as I have indicated)
in latter sections on particulate removal, stack gas conditioning,
heat recovery, separation of stack gas streams, safety require-
ments, novelty of technology, etc. - that brings into question
some of the report's unqualified conclusions. It would seem
that some of these issues should have received treatment at
least equal to the presentation on the mechanics of adsorption.
o Some recognition is given (last paragraph, page 96) to the fact
that there may be some problems involved in building and
operating these units -- and thus "pilot studies would be
necessary." The report does not indicate what specific open
issues should be resolved by such pilot studies. In general,
it is the open issues that the report does not adequately
identify and address.
cc: G. Kendall
W. Krenz
D. Lee
W. Meyer
R. Mueller
F. Thoits
M. Tyro
P. Venturini
Sincerely,
-------
RADIAN
CORPORATION
COMMENTS BY DON LEE,
VIC MANUFACTURING COMPANY
-------
VIC MANUFACTURING COMPANY
1620 CENTRAL AVENUE N E, MINNEAPOLIS. MINNESOTA 5S413/PHONE (612) 781-6601
AIR POLLUTION CONTROL SYSTEMS
March IT, 1976
Mr. Bob Wetherhold
Radian Corporation
8500 Shoal Creek 31vd
Austin, Texas 73766
RE: Carbon Adsorption WorKShop
E.P.A. REGION IX
E. P.A. Regional Office
San Francisco, CA
I. Comments on Activated Carbon - Incineration, General
Activated Carbon Adsorption is an engineering tool. It is but one of
a number of methods that an engineer must weigh to determine the best
approach on tackling any problem. Adsorption systems seem to work the
best when other methods are either impossible or too expensive. Such
is the case when a solvent is very dilute in an air stream. Condensation
becomes next to impossible and incineration gets exceedingly expensive
because it is reauirsd to heat too much air in the process. Activated
carbon adsorption is not a method of disposal of a pollutant unless,
of course, you want to take it out and bury it someplace. It should
be thought of as a method of accumulating and concentrating of a
pollutant so that it can be recycled or disposed of in some other way.
The carbon-incineration concept is merely a combination of two different
systems - one to accumulate, one to dispose of the DOllutant. The
energy value of the solvent is a bonus in these applications. The
main economy of such a system is in its ability to handle highly dilute
applications.
There has been considerable experience on har.dlin.7 baking oven exhaust.
There has been less experience on handling air from spray paint operations
In almost all of these applications, there is a limited bed life of the
activated caroon. This is because of the buildup of particulate matter,
high boiling compounds, or resins that may bake off at high temperatures.
In most cases this bed life will run anywhere from six months to six years
As such, the activated carbon system must be sized for a declining
capacity, and the activated carbon must be reactivated at sucli time as
the bed is no loncer functional. This involves some extra expense, but
not prohibitive. It should be kept in nind that activated carbon is
used in "Tertiary" sewage treatment at Lake Tahoe, wherein the activated
carbon is reactivated through a furnace on every cyc?e.
AWARDEO THE PRESIDENTS E" FOR EXCELLENCE IN EXPORT
-------
VLSBC ACAMVFACT**fftff/MC CO
Mr. Bob Wetherhold
Radian Corporation
Austin, Texas Page 2
II. Comr.ents on Ford Report
As brought out in the meeting, Ford Motor Company tested a pilot VTC
unit at tne Kdchigsn Truck Plant in Warren Michigan for approximately
2 years. This happened to be a "canister ban?." system, which is one
of several designs employed to allow a maximum air handling capacity
per system, and thereby brinf; down the total cost per CFM. Supposedly
information from such tests would be a further suide as to air handling
capacity, efficiency, etc. The pilot was provided with a tray mechanise
where different types of prefiltration nedia could be tested. It was
necessary to come up vith the proper type of media before putting it
through the carbon system to assure that it would operate properly, as such
studies were done on prefiltration media, the activated carbon syster.,
and the incineration boiler unit that was also provided with the system.
SPRAY PAINT - EXHAUST
For most of the test period involved, exhaust from spray paint booths were
tested. We were gratified to find that many types of filter media would
do a good job at this location with comparatively good life. Initial
tests on the solvent indicated that efficiencies of 65* to 95^ (typical
of Canister Tank Systems) could be obtained over an initial lb hour period.
A declining bed capacity would affect the cycle time and it could be
reduced all the way down to a minimum of 1 hour. It was felt that the
initial tests were successful.
Apparently, additional tests were done in Louisville, Kentucky under a
ouch more humid condition. This was done on the prefiltration of the
particulate matter only and carbon was not involved. From the comments
made at the meeting, it would appear that the prefiltration can be done,
but here again it's a matter of expense. The prefiltration problem is
something that would probably vary considerably from location to location.
The fact that the filter media became quite wet at Louisville would probably
indicate looking into other types of particulate filtration involving
methods other than entrainment.
Baking Oven Exhausts:
Baking oven tests were conducted later on and frankly ve never did ^et
much data back on it. However, the efficiencies that Ford experienced
are not typical, and two things remain unanswered, which are critical
to the performance of the system.
1. I doubt very much if the air was ever cooled sufficiently to do a
good Job on the solvent. The air must be cooled to temperatures below
100°F. to give us a sufficient working capacity and efficiency.
-------
*rac co.
Mr. Bob Wetherhold
Radian Corporation
Austin, Texas Page 3
The last visit I made to Warren, Michigan indicated that the type of
cooling system that vas used never pot the temperature below lllj°F.
It may be possible that lower temperatures were obtained at a later
date. It must be keut in mind to obtain a oroDer working svstem,
these temperatures must be consistent, and over a long period of time
to allow the "Adsorbed Heel" on the carbon bed to come to an
equilibrium. This won't be accomplished with one day or even several
days of lower temperatures.
2. There's no indication that the activated carbon in the system was
ever renewed nor tested prior to conducting the baking oven tests.
Since the spray paint exhaust tests were conducted for almost 2
years under a variety of conditions, there is a very good possibility
that the carbon became somewhat spent. To eliminate this as a
variable, new carbon snould have been nut into the system prior to
conducting the baking oven tests. We have many successful baking
oven exhaust applications operating presently. These are from a
variety of applications, but tne problems remain somewhat tne same —
that is, temperature and bed loading. I would say that most of these
applications were more critical from the standpoint of quantity of
high boiling compounds and loading factors. All of them have to
reactivate the carbon periodically. All of them have proven to be
less expensive than other methods of doing the same thing.
SURGING
Ford expressed a problem in that the solvent had a tendency to sura:e
during the regeneration, thereby, causin/r an uneven supply of fuel to the
incinerator boiler. This is absolutely true, although there are a number
of ways to flatten out these up and down fluctuations. This only occurred
when we were trying to incinerate the steam solvent vapors directly
without condensing. Many people prefer to condense the vapors, and then use
the liquid directly as a fuel in a boiler or burn the total liauid,water,
and solvent in a liquid incinerator. This not only allows you to eliminate
any surging effects, but it also allows you to use the liquid when and
where it makes sense. I think one of the more important things that came
out of the Ford test, as far as incinerating the vapors were concerned,
is that the incinerator apparently did a good job, whether there was a
surge or not. The company that built tne incinerator part of the equinrnent
did tests on the exhaust during the operation.
-------
MfBG MrjMMMPFjaCVMMgrMMC CO-
Mr. Bob Wetherhold
Radian Corporation
Austin, Texas Page ii
SIZE
Ford brought up the argument that there never has been any system quite
this big built, and the physical size of a proposed system was somewhat
overwhelming. They're probably right as far as overall size is concerned,
but there have been systems built handling 100,000 - 200,000 C.F.M.
There is no reason why a number of these systems wouldn't do the same
Job. In fact it's been proven on a number of different types of equipment
that once you get to a certain size, it's easier to handle the problem
with several small systems than on bir; system. There is a point of
diminishing returns when it comes to designing; one larsre system to handle
the whole thing. There are a number of arguments for a number of small
systems versus one big system:
1. The weight involved in handling various integral carts,
such as blowers, adsorbers, ducts, dampers, etc.
2. The number of smaller modules gives you considerable flexibility
during temporary shutdowns for repairs, maintenance, etc.
3. Parts are more easily replaced. Smaller modules can be built
at a factory and shipped to the job site, whereas larger systems may
have to be built at the job site at considerably more expense.
1*. Maintenance is easier on smaller systems.
It might also be pointed out, this would be a large application, no
matter what method they went to, incineration, water base systems, etc.
III. Conclusion
We also feel that the Radian Renort is very realistic. Much more
comprehensive than any similar report I have seen. It appears that the
main argument between the Radian Report and the Ford Report is that of
cost. Both seem to agree that the carbon systen does work technically.
Our experience, based on dollar per CFM, as to installation cost and
operational cost, would lean more towards the Radian Report.
The main thing to be determined here is whether Ford has to solve the
problem. At least half of our baking oven applications now existing
-------
E J8C WB4&MGJFJ& eZTTZSZSMfrSG CO
Mr. Bob Wetherhold
Radian Corporation
Austin, Texas
Pace 5
in the field were installed because of air pollution and not because of
some economic return. If Ford or anybody else has to resolve em air
pollution problem, then adsorption-incineration systems will hnve to be
considered on their own merits versus the cost of doing anythi.15 else.
At VIC we feel that activated carbon adsorption is one of the better
methods, and one of the least expensive methods. This idea is shared by
many others.
VIC MANUFACTURING COMPANY
DL:pe
cc: Ron Mueller, U.S. EPA
Chuck Gorman
------- |