v>EPA
            United States
            Environmental Protection
            Agency
            Air and Hazardous
            Materials Division
            230 South Dearborn Street
            Chicago, Illinois 60604
EPA 905/2-80-005
Determination of Capture
And Destruction Efficiencies
Of Selected Volatile Organic
Compound Control Devices
In the State of Illinois

-------
DCN 81-240-016-03-09                                    EPA 905-2-80-005



                 DETERMINATION  OF  CAPTURE AND DESTRUCTION

                         EFFICIENCIES  OF SELECTED

                 VOLATILE ORGANIC  COMPOUND CONTROL DEVICES

                         IN THE STATE  OF ILLINOIS
                                     by

                            RADIAN  CORPORATION
                            8501 Mo-Pac Boulevard
                              P. 0.  Box  9948
                            Austin,  Texas  78766
                          Contract No.  68-02-3513
                             Work Assignment  3
                               Prepared  for:
                   U. S. Environmental Protection Agency
                                 Region  V
                   Air and Hazardous Materials  Division
                            Air Programs Branch
                         Chicago,  Illinois   60604
                  EPA Project Officer:   Barry  A.  Perlmutter
                                April  1981
                       U.S. Environmental Protection Agency
                       Region 5, library (PL-12J)
                       7? West Jackson Boulevard, 12th Floor
                       Chicago, II  60604-3590

-------
                    Distribution and Disclaimer Statement
     This air pollution report is issued by Region V of the United States
Environmental Protection Agency (USEPA) to report engineering evaluations
of selected air pollution control devices to a limited number of readers.
Copies are available free of charge to grantees,  selected contractors, and
Federal employees, in limited quantities and for a nominal cost, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia, 22161.

     This report was furnished to the USEPA Region V by Radian Corporation,
8501 Mo-Pac Blvd. P. 0. Box 9948, Austin, Texas,  78766, in fulfillment of
Contract No. 68-02-3513, Work Assignment 3.  This report has been reviewed
by USEPA Region V and approved for publication.   The contents of this report
are reproduced herein as received from Radian Corporation.  The opinions,
findings, and conclusions expressed are those of the authors and are not
necessarily those of the USEPA.  Mention of company, trade, or product names
is not to be considered as an endorsement by the USEPA.
                     Publication No.  EPA 905/2-80-005
                      Lr.vlronmental Protection Agency

-------
                                 ABSTRACT

     This report provides technical support for the development of the
Illinois State Implementation Plan for surface coating industries, and more
specifically, paper and can coating.   A source testing program was conducted
at three paper coating and two can coating facilities in Illinois to de-
termine the efficiency of capture and destruction of volatile organic
compounds (VOC) using either carbon adsorption or afterburner systems.  At
the paper coaters, the VOC collection efficiencies were 91-94%, but at the
can coating plants, collection efficiency was undetermined.  On the paper
coating lines, two carbon adsorbers showed 79 and 98% control efficiency
and a thermal afterburner was performing at 95% efficiency.  The three
afterburners at the can coating plants were controlling only 26 to 73%
of the VOC's because operating temperatures were relatively low.  EPA
Method 25 was used to determine the VOC concentration in the vapor streams.
                                    111

-------
                                 CONTENTS


DISCLAIMER STATEMENT	     ii
ABSTRACT	    ill

FIGURES	    vii

TABLES	   viii

  1.  SUMMARY	      1

        Paper Coating Facilities	      1
        Can Coating Facilities	      5
        Measurement Pro cedures	      6

  2.  INTRODUCTION	      7
      2.1  Background	      8
      2.2  Test Site Selection	      9

  3.  SURFACE COATING PLANT CHARACTERIZATION	     13

      3.1  Can Coating	     14
      3.2  Paper Coating	     14

  4.  TEST PROCEDURES	     21

      4.1  Liquid Sample Procedures	     22
      4.2  Gas-Phase Measurements	     24

  5.  DISCUSSION OF RESULTS	     31

      5.1  Paper Coater, Plant C	     31
      5.2  Paper Coater, Plant D	     34
      5.3  Paper Coater, Plant E	     41
      5.4  Can Coater, Plant L	     46
      5.5  Can Coater, Plant M	     50
      5.6  Can Coaters, Plants Outside of Illinois	     58
  6.  CONCLUSIONS AND RECOMMENDATIONS	     61
      6.1  Paper Coating Operations	     62
      6.2  Can Coating Operations	     65
      6.3  Analytical Procedures for Evaluating Emission
              Control Systems	     72

  REFERENCES	     77

-------
                           CONTENTS (continued)

APPENDICES
     A.    Test Methods	      A-l
     B.    Test Instrument Calibration Data and
             Descriptive Information	      B-l
     C.    Types of Coatings and Solvents Used	      C-l
     D.    Analytical Results	      D-l
     E.    Material Balance Calculations for Plant L	      E-l
                                     VI

-------
                                  FIGURES


N timber                                                                Page

 4-1    Response of afterburner inlet VOC concentration with
          variations in coating rate at Plant C ....................    26

 •4-2    Deviation from mean of carbon analyses of replicate
          gas samples using EPA Method 25 ..........................    29

 5-1    Schematic of Plant C paper coating facility ................    32

 5-2    Schematic of Plant D coating facilities ....................    37

 5-3    Schematic of Plant D carbon adsorber system ................    40

 5-4    Schematic of Plant E paper coating facility ................    42

 5-5    Schematic of Plant L can coating facility ..................    47

 5-6    Coating line 1, Plant M ....................................    51

 5-7    Incinerator inlet VOC concentration (as CHi*) versus
          time, Plant M, Line #1 ...................................    55

 5-8    Coating line 2, Plant M ....................................    56
 5-9    Incinerator inlet VOC concentration (as CHi*) versus
          time, Plant M, Line #2 ...................................    60

 6-1    Typical measurements of VOC concentration in vicinity
          of can and paper coating lines ...........................    66

 6-2    Graphical comparison of bases for calculating emission
          control efficiencies .....................................    74
                                    vii

-------
                                   TABLES
Number                                                               Page

 1-1    Summary of VOC Collection and Recovery/Destruction
          Efficiencies of Surface Coating Operations at
1-2

2-1
2-2
3-1
3-2
3-3

3-4
3-5
3-6
4-1
4-2
5-1
5-2
5-3
5-4
5-5
5-6
5-7
Summary of VOC Collection and Destruction Efficiencies
of Surface Coating Operations at Can Manufacturers 	
Can Manufacturers with Afterburners in Illinois 	
Paper Coating Facilities with Add-On Controls in Illinois..
Characterization of Can Coating Facilities 	
Operating Parameters of Can Coating Facilities 	
Operating Parameters and Performance of Afterburners
at Surface Coating Facilities 	 	 	
Characterization of Paper Coating Facilities 	
Operating Parameters of Paper Coating Facilities 	
Operating Parameters of the Carbon Adsorber 	
Coating Characteristics: Test Values vs Plant Data........
Results of Replicate Gas Samples Using EPA Method 25 	
Solvent Usage Reported by Plant C During Test Periods 	
Plant C Operating Parameters 	
Gas-Phase Test Results for Plant C 	
Plant D Operating Parameters 	
Plant D Coatings Characteristics 	 	 	
Gas-Phase Test Results for Plant D 	
Plant E Operating Parameters 	

3
12
12
15
16

17
18
19
20
23
28
33
33
35
38
39
39
43
                                    viii

-------
                            TABLES (continued)


Number                                                                Page

 5-8    Gas-Phase Test Results for Plant E	    45

 5-9    Plant L Operating Parameters	    48

 5-10   Gas-Phase lest Results for Plant L	    49

 5-11   Plant M Operating Parameters - Line #1	    52

 5-12   Gas-Phase Test Results for Plant M - Line #1	    53

 5-13   Plant M Operating Parameters - Line #2	    57

 5-14   Gas-Phase Test Results for Plant M - Line #2	    59

 6-1    Summary of Recent Test Data at Surface  Coating
          Operations	    63

 6-2    Distribution of VOC Emissions	    70
                                     ix

-------
                                   SECTION  1
                                   SUMMARY

     The  State  of  Illinois,  in Rule 205(n)(2)(A) of its State Implementation
 Plan  (SIP), has proposed  that where operators of surface coating lines elect
 to  use  add-on controls  for volatile organic compounds  (VOC) emissions, the
 overall efficiency of capture and  destruction should be at least 75%.  This
 level of  control differs  from the  presumptive norm of  81% overall efficiency
 for add-on control equipment supported in  the U.S. Environmental Protection
 Agency's  Control Technology  Guidelines (CTG's).  Therefore, the State com-
 mitted  to complete a. study of surface coating categories currently using add-
 on  controls in  Illinois to determine what  levels of capture and control
 efficiencies are technically feasible for  a)  the current systems, and b) for
 those facilities that would  require retrofitting.

     In this study, a source testing program was conducted at three paper
 coaters and two can coating  facilities to  evaluate the current performance
 of  their  existing  add-on  control systems.  Information was collected relative
 to  possible means  of improving the performance of the control system.  Test-
 ing was conducted  on three thermal afterburners, one catalytic afterburner,
 and on  two carbon  adsorbers.  To prevent disclosure of proprietary informa-
 tion, the facilities tested were assigned  code letters.  A summary of the
 findings at the five test sites is presented in Tables 1-1 and 1-2 for the
 paper and can coaters, respectively.

     Paper Coating Facilities

     At two of the paper coaters,  C and E,  the  collection  efficiencies
were 94-98% and 94%,  respectively,  but  it was undetermined  at  D.  The thermal
oxidizer at C  was operated at 95-96%  destruction efficiency giving a  91-93%
overall control.  The carbon adsorber at  E was  determined  to have a 98%

-------
            TABLE 1-1.  SUMMARY OF VOC COLLECTION AND RECOVERY/DESTRUCTION EFFICIENCIES

                        OF SURFACE COATING OPERATIONS AT PAPER COATERS
Plant Code

Collection Efficiency, %
(Ei)
Afterburner (Destruction)
or Carbon Adsorption
(Recovery) Efficiency, %
(E2)
Overall Control
Efficiency, % (E3)
(E3 = EI x E2)
Source of Data
Source of Data

A l

-

3
96-97



-

Ref 5


B 2

-

I*
-


5
90

Ref 6


C

94-98

3
95-96



91-93

Field
Test
10/80
D

Unk

•»,6
79



<79

Field
Test
10/80
E 2
7
94

",6
98


5
92

Field
Test
10/80
F

-

3
85



—

Ref 7


G 2

-

i*
-


5
93-96

Ref 8


H '

-

3
97-99



-

Ref 9


I 1

-

3
98-99



-

Ref 10


Remarks:
 l
   Gas analyses by flame ionization detector.
 2
   Solvent balance.
 3
   Thermal oxidizer.
 "t
   Carbon adsorber.

 5 Based on makeup solvent requirements and data supplied by the plant.

 6 Based on VOC measurements of adsorber offgas.

 7 Calculated: EI = E3 ^E2

-------
                      TABLE 1-2.  SUMMARY OF VOC COLLECTION AND DESTRUCTION EFFICIENCIES
                                  OF SURFACE COATING OPERATIONS AT CAN MANUFACTURERS
Plant Code
Type of Operation
Type of Afterburner
Collection
Efficiency, % (Ei)2
Afterburner
Destruction
Efficiency, % (E2)
Overall
Control
Efficiency1 % (E3)
(E3 = E! x E2)
Source of Data
Source of Data
L
Sheet
Coating
Thermal
2212'"
73
-
Field
Test
10/80
M
Sheet
Coating
Line 1
Thermal
174"
26 5
-
Field
Test
10/80
Line 2
Catalytic
903
49 5
44
Field
Test
10/80
P
Inside
Spray
Thermal
77
_
-
Ref 2

Q
Sheet
Coating
-
74-79
.
-
Ref 3

R
Sheet
Coating
Thermal
73
89
65
Ref 4

S
Inside
Spray
Thermal
12
_
-
Ref 4

u>
     Remarks:

        Assuming all  captured emissions are routed to afterburner.
      f)
        37%  oE  the  "captured" vapors are subsequently vented directly to the
        atmosphere  from  the oven's cooling zones rather than being sent to  the
        afterburner.
      3 12%  of  the  "captured" vapors are subsequently vented directly to the
        atmosphere  from  the oven's cooling zones rather than being sent to  the
        afterburner.
      '' Unexplainable high value based on plant sampling  and production data
        supplied by plant.
      5 Low  efficiency attributed to low operating temperature  of afterburner.

-------
recovery and a 92% overall efficiency.   The carbon adsorption system at
Plant D was performing at only 76% efficiency.   No explanation for the low
value was available but the normal design performance is in excess of 90%.
Results of control system performance at six other facilities in the country
are included in Table 1 to supplement the limited testing of this program.
At three plants, A, H, and I,  the thermal oxidizers (afterburners) performed
at 96-99% efficiency, but at Plant F, which operated its oxidizer at a
lower temperature, only 85% was observed.  The  overall control efficiencies
of 90 and 93+% at B and G, respectively, reflects high collection and carbon
adsorption performance.

     From these observations,  it may be concluded that a collection
efficiency of 90+% at any given paper coating facility is reasonable if:

        the coating applicator zone is  close-coupled or hooded to the
        drying oven,
     •  the coating room is sealed from the outside and maintained under
        a slight negative pressure so that essentially all the VOC
        emissions from the coater are drawn into the oven,  and
     •  all exhausts  from the  oven are ducted to an add-on control
        device, i.e., a carbon adsorption bed or an afterburner.
     Carbon bed adsorbers can be designed to provide at least 90% re-
covery of VOC  emissions introduced into the control system.  Continuous
operation at close to design performance is possible if the system is
properly monitored and maintained.  Similarly,  thermal afterburners have
been demonstrated to destroy at least 90% of the VOC emissions entering the
system, especially if operated at a temperature in excess of 1400°F and
with an adequate residence time, e.g.,  0.3 seconds.

-------
      Therefore,  at  paper  coating  facilities,  overall  control  of VOC
 emissions  at  81% or higher  efficiency  appears  technically  feasible on  the
 basis of achieving  90%  capture  of the  emissions  and 90%  control of the
 captured emissions  using  a  conventional  add-on control system.

      Can Coating Facilities

      At the two  can coating  facilities tested  in this program, Plants  L
 and M, sheet  base coating was the only operation controlled as shown in
 Table 1-2.  On one  coating line,  a VOC collection efficiency  of 90% was
 observed but  at  two other lines,  the data imply  efficiencies  in excess of
 100%  and must be discarded.  Because the control systems are  not operated
 in the winter months, it  was not  possible to repeat the  testing.  To
 supplement these observations,  results were included from  four facilities
 in other states.  The VOC collection efficiencies of the sheet coating
 operations at Q  and R sites were  73 to 79%.  For sites P and  S, where  an
 inside spray  operation was performed, the collection efficiencies were 77
 and 12%, respectively.  Earlier studies have shown that  the distribution of
 emissions between the coater-flashoff area and the drying  oven varies with
 the type of can  coating operation being performed.   When base coating
 sheets, only  10% of the total emissions "may occur at the coater-flashoff
 areas, but with  inside spray operations,  up to 80% of the  solvent flashes
 off.  The balance of the  emissions come from the oven.  Therefore, the low
 value  reported for  Plant  S is understandable,  and reflects the range in
 performance that may be observed at different  coating facilities.
     Once the coating material enters the drying oven, the potential
VOC emissions are contained.  However, the oven itself is a source of emis-
sions if it is not maintained at a negative pressure relative to its surround-
ings.  Also, the cooling zones of the oven may vent VOC's unless the zones
are exhausted to an add-on control device.  At Plants L and M,  the cooling
zones of the drying ovens exhausted directly to the atmosphere and contained
37% and 12%, respectively, of the VOC vapors leaving the ovens.

-------
     To obtain 90% or better collection efficiency, one alternative is to
completely'enclose the existing sheet coating lines as well as all other
coating operations.  This is especially true for plants which operate
several coating operations in one building.   However, the technical feasi-
bility of maintaining a proper air balance between the room and the oven
zones is not readily determined.   An alternate method of enhancing collection
efficiency is through the use of more hooding and floor sweeps.  In any case,
the oven cooling zones should be designed and operated to minimize exhausting
VOC's to the atmosphere.

     The afterburners tested at the can coating plants in this program
were performing at relatively low levels, i.e., 26-89%.  In some cases, the
systems were designed for low temperatures primarily for odor control and
hence their performance may be improved by increasing the operating
temperature.

     In summary, the level of overall control of emissions from any
given can  coating operation is heavily dependent on the technical feasibility
of enhancing the capture efficiency of VOC's from the coating line.  Can
coating lines are frequently in large, unsealed buildings and may be laid
out in a more open fashion than paper coating lines.

     Measurement Procedures

     The total gaseous nonmethane (TGNMO) procedure, EPA Method 25, was
selected for the gas-phase analyses in lieu of using a flame ionization
detection  (FID) technique to determine the carbon concentration of streams
to and from add-on controls.  The sampling procedure is relatively simple
and the laboratory analytical procedures have been demonstrated.   When
analyzing mixtures of solvents and/or products of partial combustion, it
may be more cost-effective than using gas chromatographic-mass spectrometric
techniques.  A need was identified to develop a modified Method 25 procedure
to analyze for the concentration of carbon in the liquid coating materials
to more readily determine the capture efficiency of coating lines.

-------
                                 SECTION 2
                                INTRODUCTION

     In the proposed Illinois State Implementation Plan (SIP), Rule 205(n)
(2)(a) contains an alternative compliance option for owners and operators of
surface coating lines whereby the sources are exempt from the requirement to
use high solid coating materials if emissions of volatile organic compounds
(VOC) are controlled by an afterburner system.  The overall efficiency of
capture and destruction of the nonmethane volatile organic materials is pro-
posed to be at least 75%.  This differs from the 90% capture of emissions or
the 81% overall control efficiency (capture plus control) which represents
the presumptive norm for add-on control equipment as technically supported
                                            11
in the Control Technology Guidelines (CTGs).  Because the State did not sub-
mit technical support for the lower level of control, the State committed to
complete a study to evaluate the VOC categories currently using add-on con-
trol technology (i.e., afterburner systems) in Illinois and to determine the
capture and destruction efficiencies for these systems along with associated
plant engineering data.  This report describes the details and observations
of a field sampling study performed during October, 1980 as a part of this
committment.

     The U.S. EPA in its final determination, 45FR11482, stated that,
although 90% capture systems are Reasonably Available Control Technology
(RACT) as a general matter, because the CTGs do not address unique circum-
stances, the State is free to determine RACT on an individual source basis.

-------
 2.1  BACKGROUND

     The surface coating industry consists of eight major categories of
 sources:

     •  Cans
        Paper
     •  Fabric products
     •  Metal coils
     •  Magnet wire
     •  Metal furniture
     •  Large appliances
     •  Automobile and light-duty trucks

     Although the coating materials, process equipment, operating techniques
and control technology used by one source may also be used in part by another
source in a different category, there are considerable variations or options
used by the industry as a whole as well as unique circumstances for many
facilities.
     Based on earlier reviews,  the EPA Regional Office V requested that the
emphasis be placed primarily on the can and paper coaters.   In two categories,
metal coils and magnet wire, the VOC emissions are currently effectively con-
trolled by afterburners and/or  use of waterborne formulations.  Control
technology used by the can and  paper coaters may possibly be transferred to
the two major fabric coating facilities in Illinois.   In the other three
categories, automobile/truck, metal furniture, and large appliance coating,
it is anticipated that implementation of high solids  and/or waterborne
formulations,  as well as improved transfer efficiencies,  will  occur  rather
than the continued use of solvent-based formulations  and add-on  controls.

-------
     Although the study did not address the overall compliance plans
for individual facilities, it should be noted that the Illinois SIP,
in Rule 205(n)(4), provides for the use of internal offsets.   This allows
a plant to apply the "bubble concept" and use a combination of control tech-
niques.  Thus, for those specific operations in which it is difficult to
capture a high percentage of the VOC emissions in a concentration that can
be incinerated or adsorbed, a company may elect to use the bubble approach
in their compliance plans.

2.2  TEST SITE SELECTION

     The general criteria for selecting can and paper coating
facilities for field testing included:

        located in the State of Illinois, preferably near Chicago,
     •  currently using add-on controls, i.e., afterburner or
        carbon adsorption system,
     •  operation representative of other facilities in the state, and
        employs control technology that can be applied to other
        surface coating facilities.

     Although facilities with afterburners were of more concern because they
are mentioned specifically irt Rule 205(n)(2)(A), carbon adsorption systems
were also investigated because of their applicability and ease of sampling.

     Other desired, but not necessarily required criteria, were those
specific to the technical aspects such as:

        the coating formulation uses a single solvent or a mix that
        can be analyzed with reasonable accuracy by a flame ionization
        detector  (FID)
        solvent consumption can be readily measured

-------
         sample points exist or can be easily installed on the control
         device inlet and outlet
         solvent-laden air flow in and out of the control device can
         be readily measured
         each control device services only one coating line

     The major can and paper coating facilities in the Chicago area were
first reviewed on the basis of information obtained from the State of Illi-
nois and then subsequently by telephone calls.   Finally, the most likely
candidate sites were visited prior to testing to determine the feasibility
of obtaining adequate performance data.

     To preclude disclosure of proprietary information, all of the sources
have been assigned code letters and will be referred to by these letters in
this report.

2.2.1  Can Coating

     Of  the 21 major can manufacturing facilities in Illinois, only 10 are
currently using afterburners (A/B).   Six plants have thermal units, two have
catalytic, and two have both types.   These controls are generally on the sheet
coating ovens, but at two plants, they are on the lithographic press lines.

     The total number of emission sources  at  can coating facilities in the
state is approximately 434.   Of these,  only  73  are  coating ovens  and 51 are
press ovens.   The distribution of emission sources  by types  of operations and
number of sources controlled follows:
                                     10

-------
                                      Number
                    Total Units     Controlled     Type of Control
Coating Ovens            73              36         24 Thermal A/B
                                                   12 Catalytic A/B
Press Ovens              51              10          8-Thermal A/B
                                                    2 Catalytic A/B
All Other Operations*  310               4          3 UV Resin
	                                               1 Heat Sensitive Coating
*Includes inside and outside stripe, side seam, inside coating, end coating,
 end sealing compound, and body spray.

     The type of controls used by ten can manufacturing plants is
presented in Table  2-1.  All but one of the sources are in the Chicago
area.

     Two sources, L and M, were selected for field sampling on the basis of
their meeting most of the major criteria.   In no case does any source use a
single solvent or simple mixture of compounds that would allow ready analysis
of the emissions using a flame ionization detector (FID).  For that reason,
it was decided to use the recently developed EPA Method 25 to determine
total gaseous nonmethane organic carbon concentration as described in
Section 4.

2.2.2  Paper Coaters

     Of the 31 major paper coating facilities  in Illinois,  only six are using
add-on controls,  three of which have afterburners and the other three use
carbon adsorption.   (Two plants use  waterborne coatings  and one uses  ultraviolet
curing.)   The type of  control used in these plants  is  given in  Table  2-2.

     Any source using  a carbon adsorption  system can normally provide reliable
data on capture efficiency since the makeup solvent rate to the system, less
any liquid withdrawals to controlled waste solvent disposal and VOC left in
the absorber off-gas,  equals the fugitive  VOCs not captured.   Source  D and  E
using carbon adsorption and  Source C using afterburners  were  selected as can-
didates for field testing.   As with  the can coaters,  none of  the paper coaters
uses a "simple" solvent mixture in its formulations.
                                     11

-------
        TABLE 2-1.  CAN MANUFACTURERS WITH AFTERBURNERS IN ILLINOIS
Facility Code
Operation Controlled
Type of A/B Control
L
M
N
T
U
V
W
X
Y
Z
( ) ** Number of
* = One contrc
Coating Oven
Coating Oven
Coating Oven
Litho Oven
Coating Oven
Lithographic Oven
Coating Oven
Coating Oven
Coating Oven
Coating Oven
Coating Oven
Coating Oven
(3)
(8)
(2)
(6)
(2)
(3)
(4)
(6)
(9)
(4)
(8)
(6)
Thermal
Thermal
Catalytic
Thermal
Thermal
Catalytic
Thermal
Thermal
Thermal
Catalytic
Catalytic
Catalytic
Thermal
Catalytic
(3)
(3)
(5)
(2)
(2)
(1)
(2)
(3)
(4)
(2)
(D*
(4)
(8)
(2)
ovens or control devices
3! on 3 ovens
    TABLE 2-2.   PAPER COATING FACILITIES WITH ADD-ON CONTROLS IN ILLINOIS
Facility Code
A
C
D
E
J
K

Number
1
7
6
3
7
3
Controls
Type
Afterburner
Afterburner
Carbon Adsorber
Carbon Adsorber
Carbon Adsorber
Afterburner
Control Device
Efficiency*
%
97
98
97
90
85
97
'•Reported by plant management.
                                      12

-------
                                 SECTION 3
                  SURFACE COATING PLANT CHARACTERIZATION

     The five surface coating plants selected for field testing to
determine VOC emission capture efficiency and afterburner  or  carbon ad-
sorption efficiency,  while typical of other such plants in Illinois,  do have
some unique features  that are cited in the following paragraphs.  In certain
cases, they affect the overall performance value and,  therefore,  these factors
should be considered  in determining what is optimum performance for a normal
installation.

     The VOC emissions from a paper or can coating line come  either from
the application-flashoff area or from the drying ovens. For  can coating
operations, there may also be emissions associated with the cooling zone.
(Depending on the coating operation being performed, the application-
flashoff emissions may range from 0 to 100%,  and 0 to  100% may come from the
ovens.)  Although VOC vapors released in the drying ovens  are normally con-
tained at close to 100% efficiency (for safety and health  reasons), they may
either be exhausted to the atmosphere or be processed  by add-on control
systems, i.e.,  carbon adsorption or incineration.  Where a mixture  of solvents
is used, or where the organic material cannot be recycled, incineration is
generally the preferred control.

     The design parameters for incinerators are well established and
systems are readily designed to perform with 90-95%, or higher, destruction
of VOC.1  The operating efficiency of afterburners, both thermal and
catalytic, can be determined by analyzing the inlet and outlet streams.  The
primary design parameters are oxidizer temperature and residence time.
                                     13

-------
     The efficiency of capture of emissions from the coating equipment
and flashoff area is difficult to measure unless the entire coating line
is enclosed and operated under negative pressure.   One method to determine
capture efficiency is by a material balance wherein the fugitive loss is
based on the difference between two large values,  i.e., the mass of solvent
input to the coater, less the mass of solvent ducted to the add-on con-
trol device.

3.1  CAN COAT INC?

     A general characterization of the selected plants, "L" and "M",  is
presented in Table 3-1 and includes the process operations performed, the
types of vapor collection and afterburners used.  Table 3-2 indicates the
operating parameters of the coating lines as reported during the pretest
site visits.  The types of solvents and coating materials used are listed
in Appendix C, but no specific formulations are indicated since they  are
proprietary and also are varied quite frequently.   The typical operating
parameters of the afterburners are shown in Table  3-3.  In Tables 3-1, 3-2,
and 3-3, data are also included for facilities not tested in this study but
for which performance data are available from earlier investigations  by
others.

3.2  PAPER COATING

     Three paper coating plants, "C", "D", and "E",  were tested during
this project.  The general characteristics of these plants are presented
in Table 3-4 and the coating line operating parameters in Table 3-5.   The
afterburner operating parameters are shown in Table 3-3, and the carbon
bed parameters in Table 3-6.  As was the case with can coating, the specific
formulations used in the coating are proprietary so only the components are
indicated in Appendix C.
                                    14

-------
                    TABLE 3-1.  CHARACTERIZATION OF CAN COATING FACILITIES
Plant
Code
L
M

P
Q
R
S
Products
Sheet Coating
(Enamel Plate)
Two-Piece and
Three-Piece Cans

Two-Piece Cans
Three-Piece Cans
Three-Piece Cans
Three-Piece Cans
Processes
Sheet Base Coating
(Exterior and
Interior)
Sheet Base Coating
Printing/Lacquer
Coating
Interior Spray
Coating
Sheet Base
Coating
Sheet Coating
Interior Spray
No. of SLA Collection Afterburners
Lines System* No. Type
2 Hoods over 2 Thermal
Coaters
8 None 5 Catalytic
3 Thermal
9 - 9 Thermal
4 Hood - Thermal
1 None None -
6 - - Thermal
9 - - Thermal
Remarks
(1)
(1)
-
(2)
One.
Cooling
Zone
(3)
(4)
(4)
*In addition to oven exhaust.   SLA =  Solven-laden air.

Remarks:  (1)  Tested 10/80.
          (2)  Tested 10/79.  See Ref. 2.
          (3)  Tested 10/79.  See Ref. 3.
          (4)  GARB Study, Ref. 4.

-------
                  TABLE 3-2.  OPERATING PARAMETERS OF CAN COATING FACILITIES
Plant
Code
L
M

P
Q
R*
s*
Line
Number
1
1
2
4
1
1

Line Speed
(sheets/min)
Range Average
85-105 105
90-110 100
-
650 cans/
min
100 max
75
380 cans/
min
Type of
Coating
Enamels
Vinyls or
Epoxies
Print Ink
Inside Spray
Base Coat
-
"
Coating Weight
(rag/ in2) (wet)
Range Average
3.5-10
1-12 11.0 Vinyl
4.5 Epoxy
Varies 3.0
-
-
4.2
0.28 g/can
Run
Length**
(hrs)
72-120
24
>24
-
-
-
"
*See Reference 4.
**Supplied by plant management.

-------
                         TABLE  3-3.   OPERATING  PARAMETERS  AND  PERFORMANCE OF
                                        AFTERBURNERS AT SURFACE COATING  FACILITIES
Plane
Code
Unit
No.
Type of
Afterburner
Can Coaters
L

M

Paper
c2

1
2
1
2
Coaters
1
1
Thermal
Thermal
Thermal
Catalytic
Thermal
Thermal
Heat exchange
Primary

Yes
Yes
No
No
Yes
Yea
Secondary

No
Yes
No
No
No'
No3
Auxiliary fuel SLA
Tvoe 10"Btu/hr MSCFM

NG 13
NG
NG
NG
NG
NG

-
8
4.4
7.0
17.9
17.9
Flow
"F

350-400
350-450
148
228
342
342
Outlet Flow
MSCFM

25-35
25-35
3.2
6.4
12.5
12.5
°F

1300-1500
1300-1500
551
548
1250(870)
1250(870)
VOC (Ib/hr) Afterburner
In


0.
0.
0.
7.
7.

-
348
349
566
89
54
Out Eff.. %

-
0.0963
0.260
0.291
0.43
0.27

-
72
26"
49"
95
96
Data source

Plant Mgr.
Test 10/80
Plant Mgr.
Plant Mgr.
Test 10/80
Test 10/80
 Measured us carbon.
2Plant C was te&tud during two separate periods.
3Aftt_Tburner exhaust directly reused in drying oven.
NG = Natural gas
SLA= Sol vent-ldiitm air
""Low efficiency attributed to low operating temperature .

-------
                            TABLE  3-4.  CHARACTERIZATION OF  PAPER COATING  FACILITIES
00
Web
Code Material
C Paper
D PVC film
Mylar film
E Paper
No. of Continuity
Lines of Operations
4 24 hr/day
6 day/wk
2 24 hr/day
5 day/wk
2 24 hr/day
5 day/wk
Condition SL#< Collector
of System System
Very Good Oven exhaust only.
Applicator close
connection to oven
New Complete coating
room control
Very Good Complete coating
room control
Control
Device
Thermal
afterburners
Carbon
adsorber
Carbon
adsorber
       *SLA=Solven-laden air

-------
                    TABLE 3-5.  OPERATING PARAMETERS OF PAPER COATING FACILITIES
Plant
Code
C
D





E


Line
Process No.
Paper Coating 1
Web Coating 1


2


Paper Coating 1

2
Line Speed
f t/min
(Proprietary
45


55.5


60

60
SLA Control Device
Type of Coating SCFM °F
Information) 18,000 1250*
Pressure-Sensitive
Adhesive & Release

Pressure-Sensitive
Adhesive, Laminate
Adhesive & Release
Nitrocellulose

Nitrocellulose
702

30,000




10,400 7Q2

 Thermal afterburner
"Carbon Adsorber

-------
            TABLE 3-6.  OPERATING  PARAMETERS OF THE CARBON ADSORBER

Plant
Code
D
E

Unit
No.
1
1

Number of
Beds
3
3

Pounds of Carbon
Per Bed
7,000
6,000
Superficial
Velocity
(fpm)
<100
90
Pounds of Steam
Per Pound of
Recovered Solvent
NA
5-6

SLA
MSCFM
30,000
10,400

Flow
°F
72
90
NA - Not Available

-------
                                 SECTION 4
                              TEST PROCEDURES

     Test procedures were selected to permit the determination of the
efficiency of any VOC collection system used to recover a mixture of known
or unknown solvents from a coating line and to measure the efficiency of the
add-on controls.  The procedures were developed either by EPA or the
American Society for Testing and Materials (ASTM) or are slight modifications
thereof.

     Because the surface coaters studied in this program do not use a
single solvent but rather use a mixture of solvents, the analysis of gas
streams by conventional gas chromatography or flame ionization detector is
very difficult or nearly impossible.  Therefore, it was necessary to use a
carbon balance approach to determine the efficiencies.  EPA Method 25 was
selected to provide the solvent concentration in the gas streams since the
result can be expressed in moles of carbon per volume of gas and is
independent of the molecular configuration of the solvents present.  Using
this method obviates the problems associated with trying to identify the
actual solvent species or their derivatives resulting from partial oxidation
in ovens or afterburners.  The need to calibrate any FID for each VOC
species in the concentration range that exists in the plant gas stream is
also avoided.

     Capture efficiency was determined by comparing the mass of carbon
passing through the collection system per unit time (using Method 25) with
the mass rate of carbon charged to the coating applicator via solvents.
The mass rate was determined from the measured coating rate, analyses of
the density and volatile content, and the carbon content of the solvent
fraction.  In one case, the mass rate was determined by the plant personnel
to maintain the proprietary formulation.

                                    21

-------
     The efficiency of the control device was determined by comparing the
amount of carbon recovered (in carbon adsorbers)  or destroyed (in after-
burners) to the mass of carbon entering the control device.  For after-
burners, the efficiency is determined through gas-phase analysis for
carbon content.  For carbon adsorbers, gas-phase analysis and solvent
balances were used.

     The procedures for sampling and analyzing the liquid and gas streams
are discussed in the following paragraphs.

4.1  LIQUID SAMPLE PROCEDURES

     Grab samples were taken from each coating line for the different solvent-
based coatings applied during the test runs.  The samples were analyzed in the
Radian laboratory for volatile content using ASTM Procedure D2369-80 and for
density using ASTM D1475-60.  Portions of the coatings were distilled using
ASTM Procedure D3272-76 to provide a solvent that could be analyzed for car-
bon content using a modification of EPA Method 25.  This latter analysis was
performed by Truesdail Laboratories, Inc., Los Angeles, California.  From
these four analyses and procedures, it was possible to calculate the mass of
carbon per mass of coating charged to the applicator.   Details of these test
methods are provided in Appendix A.

     The results of the-individual laboratory analyses for density,
volatile content, and mass of carbon per mass solvent  are compared with the
estimates provided by the plant personnel in Table 4-1.  Not all of the
plants supplied complete data on their coatings,  but the table identifies
those plants where data exists.   In general, the  laboratory analyses of the
coating samples are supported by the estimates provided by the plant.

     The coating application rate was estimated from plant process data.
The most reasonable approach was to take the coating weight data determined
routinely by the operators and multiply that by the process feed rate.   The
                                    22

-------
TABLE 4-1.   COATING CHARACTERISTICS:   TEST VALUES VS PLANT DATA
Plant
E
E
L
L
D
1,'A = Not Av<
Type of Coating
Ink
Ink
Enamel
Enamel
Adhesives
•liable
Test
0.
0.
1.
1.
Density
Value
8894
8858
0355
3744
(gm/cm3)
Plant Estimate
NA
NA
1.013 + 0.05
NA
Test
71.
66.
22.
34.
Volatile
Value
1
6
7
1
Content, HtX
Plant Estimate
68
68
18-25
NA
Mass C /Mas
Test Value
0.
0.
0.
0.
748
748
848
725
.87-. 90








;s Solvent (Ib/lb)
Plant Estimate
0
0

0
0

.748
.748
NA
.735
.88


-------
coating weights for can coating operations are generally expressed in milli-
grams per one square inch or four square inches, while the paper coaters
generally express their weights in pounds per 3000 square feet (one ream).
The feed rate for can coating is usually in sheets per minute while paper
coating, which is a continuous web operation, is in feet per minute.

     Other methods for calculating the coating application rate were
considered but rejected.  Measuring weight or volumetric changes in the
coating feed tanks was not feasible primarily due to the inaccuracy of tank
measurements taken over a relatively short (one hour)  test period, and the
lack of accurate volumetric data on the applicator tank.  Also, in some
cases, minor modifications are made to the formulations during the run by
adding solvents or thinners.  It is very difficult to  maintain an accurate
record of these additions.

4.2  GAS-PHASE MEASUREMENTS

     The gas-phase measurements consisted of volumetric flow,  molecular
weight, moisture,  and VOC concentration.  These measurements were performed
as described in EPA Methods 1 through 4 and 25.  (See  Appendix A)  The gas
analyses were performed by Pollution Control Science,  Inc.,  (PCS) of
Miamisburg,  Ohio,  and by Truesdail Laboratories, Inc., of Los Angeles,
California,  using Method 25.  The velocity, molecular  weight,  and moisture
analyses were performed by the on-site Radian test team.

     An S-type pitot tube and a magnehelix were used to obtain the pressure
readings for the velocity measurements.  The pitot and magnehelix gauges were
calibrated for the tests and their calibration curves  are given in Appendix
B.  At one plant,  an Alnor® velometer had to be used to measure the duct gas
flow rate.  Information on the Alnor and its calibration is  also provided
in Appendix B.
                                    24

-------
     With Method 25, the bulk of the solvent vapors in the gas stream is con-
densed in a cold trap chilled with dry ice.   Any noncondensible gases includ-
ing noncondensible organic vapors pass through the trap and are collected in
an evacuated cylinder of known volume.  Both the trap and evacuated cylinder are
then analyzed off site for total gaseous nonmethane organics.   These values
permit the calculation of the total carbon concentration in the gas stream
expressed in parts of carbon per million of  gas on a volume basis.   A detailed
description of Method 25 is given in Appendix A.

     To supplement the Method 25 tests, the  solvent concentrations  were mea-
sured by a Century® Organic Vapor Analyzer (OVA)  and by charcoal tube grab
samples.  The OVA was used to monitor the system performance.   Usually, the
sample probe was located at the control device entrance and the monitor
showed the variations in solvent concentration due to process  upsets, start-
ups , and shutdowns.

     The charcoal tube samples were taken as a backup to the Method 25
analysis.  A measured amount of solvent-laden air was passed through the
tube over about a 30-minute period and the activated charcoal  adsorbed the
VOC emissions.  The tubes can be analyzed in an off-site laboratory to
determine the solvent concentration.  Due to limited resources and  the fact
that the gas analyses by Method 25 appeared  to be satisfactory, the charcoal
tubes were not analyzed.

     An indication of the accuracy of the Method 25 tests can  be seen from
the results of the tests at Plant C.  Plant  C personnel provided the rate of
solvent applied with the coating in terms of carbon mass per minute for each
of three sample periods.  For these sample periods, a different solvent
application rate was used.  The average VOC  concentrations measured during
these periods using Method 25 correlated well with the VOC application rates
calculated by the plant and based on their proprietary formulation
(Figure 4-1).
                                   25

-------
•O
O
o
ja
     15,000
 a
 P.
14,500
 M
 (U

 C
 M



 I

 0)
-S

 C
 o
 CO
 •t-i
 4-1

 s
 o

 o
 8
14,000
      13,500
                                    JL
                                           _L
J.
                      7.8          7.9           8.0         8.1


                                Solvent  Application Rate, Ib C/min
                                                                   8.2
                  Figure  4-1.   Response of afterburner inlet VOC concentration

                                with variations in coating rate at Plant C.
                                       26

-------
     There were not enough analyses performed to determine the
precision of the data when using EPA Method 25.   However,  repeatability
over a concentration range from 200 to 15,000 ppm is indicated in
Table 4-2 and in Figure 4-2.   With two exceptions,  the range of analyses
is 275 ppm or less and is relatively independent of the concentration.
                                   27

-------
                    TABLE 4-2.  RESULTS OF REPLICATE GAS SAMPLES USING EPA METHOD 25
Percent
Carbon Concentration, ppm Deviation From
Plant ID
C
C
L
L
L
L
M
(Line 1)
M
(Line 2)
Samples ID
Afterburner Inlet
Afterburner Inlet
Afterburner Inlet
Afterburner Outlet
Cooling Zone A
Cooling Zone B
Afterburner Inlet
Afterburner Inlet
Test 1
15,290
13,895
1,696
613
259
519
2,320
2,783
Test 2
14,431
13,620
1,527
454
180
263
3,146
2,688
Mean
14,860.5
13,757.5
1,611.5
533.5
219.5
391
2,733
2,735.5
the Mean*
+ 2.9%
+ 1.0%
+ 5.2%
+14.9%
+18.0%
+32.7%
+15.1%
+ 1.7%
Range
of Analyses
ppm
859
275
169
159
79
256
826
95

*Percent deviation from the mean DFM calculated as follows:
     = (Test 1 - Test 2)/2
       (Test 1 + Test 2)/2
100

-------
         RESULTS of  REPLICATE  GAS SAMPLES USING  EPA METHOD 25
E 35
R
C
E
N
T

D 25
E
V
I
A
T
I
0
N

F
R
0
M
M
E
A
N
      a
   15
   10
0
   a
            a
                                                             a
                                                         a
     0
         2500     5000     7500     10000     12500
                 MEAN  CARBON CONCENTRATION—ppm
                                                                15000
              Figure 4-2.  Deviation from mean of carbon analyses of
                         replicate gas samples using EPA Method 25.

-------
                                 SECTION 5
                           DISCUSSION OF RESULTS

     Field tests were conducted at three paper coating and two can coating
facilities in Illinois to measure the actual efficiency of VOC capture or
collection systems and the efficiency of add-on control devices to recover
or destroy the VOC emissions.  In addition, engineering data were collected
at the plant sites to assist in the estimation of reasonable performance
for any plant with a well-designed and operated VOC collection and abate-
ment device.

     The raw data collected in the field are assembled in Appendix D.  These
data are used to calculate the application rates and gas flow in the ducts
leading to and from the control devices.  The results of the liquid and
gas-phase analyses are also provided in Appendix D.  Sample calculations
are included in Appendix E.

     This section presents and discusses the results of the material
balance calculations for collection,  control device, and overall efficiencies
at the sites tested.

5.1  PAPER COATER,  PLANT C

     The paper coating facility tested at Site C consists of a single
coating line controlled by a thermal  oxidizer-type afterburner.   A schematic
of the facility is shown in Figure 5-1.    Due to the confidentiality agreement
with Plant C, the test contractor was not allowed in the coater  room.  Also,
the contractor could not take grab samples of the coatings or any engineering
data on the coating line.   However,  Plant C did provide the solvent applica-
tion rate expressed as total carbon (Table 5-1) and some coating line and
afterburner operating parameters (Table 5-2).

                                   31

-------
OJ
ro
                                                                                            Afterburner  Outlet
                                                                                            Sample Point
                                                 Afterburner Inlet
                                               VOC Measurement Point
Afterburner
Exhaust Stack
                                                                                           Afterburner
                                                                                           Inlet Velocity
                                                                                           Measurement Points
      Makeup Air
     	Trom
      Outside The
      Building
                                      Figure  5-1.   Schematic of  Plant C  paper coating facility.

-------
               TABLE 5-1.  SOLVENT USAGE REPORTED BY PLANT C
                           DURING TEST PERIODS

                        Test Date:  October 29, 1980
Time
1200 -
1335 -
1600 -
1810 -
1335
1600
1810
2040
Carbon
10.9
11.0
10.7
10.9
atoms /hr
x 1027
x 1027
x 1027
x 1027
Ib Ci /Minute
7
8
7
7
.988
.061
.841
.988
Gas sampling was performed during the time intervals of 1435 - 1535,
1655 - 1755, and 1917 - 2017.
                  TABLE 5-2.  PLANT C OPERATING PARAMETERS
Coating Line Operating Parameters

     Web substrate           -  continuous paper roll

     Line speed              -  (confidential)

     Line width              -  (confidential)

     Coating material        -  adhesive and release

     Coating weight          -  (confidential)

     Weight percent solvent  -  (confidential)

     Solvent composition     -  (confidential)

Afterburner Operating Parameters

     Type of Afterburner               -   Thermal oxidizer

     Solvent-laden air to afterburner  -   18,000 scfm

     Afterburner temperature           -   1,250 °F

     Residence time                        0.3 seconds
                                     33

-------
     Gas-phase analyses were performed at the afterburner inlet and outlet.
The results of these measurements are shown in Table 5-3.  There were no
accessible sample locations directly before the afterburner unit.  There-
fore, the inlet flow rate had to be estimated from the sum of the combined
outlet flow rates of the three hot air streams leaving the drying/curing
oven.  Due to physical limitations in the location of the sample ports in
the oven outlet ducts, the S-type pitot could not be used to measure the
velocity in the ducts.  Instead, an Alnor velometer had to be used.
(A discussion of the Alnor velometer is given in Appendix A.)

     The expected VOC capture efficiency for this facility is 100 percent
based on the assumption that all VOC-laden gas streams (or SLA) are even-
tually pulled through the drying/curing oven and then through the afterburner.
The VOC capture efficiency was measured for two distinct one-hour time
periods.  From 1435 to 1535, the solvent capture efficiency was 97.9% of
the solvent applied at the coater.  For the period 1920 to 2020, the
capture efficiency was 94.4%.

     The VOC destruction across the afterburner was determined to be
about 95 to 96%.  The temperature in the incinerator was 1250°F and the
residence time was estimated at 0.3 seconds.   This good performance of the
afterburner may be attributed in part to the high VOC concentration in the
inlet stream.

     The overall VOC control based on the capture and destruction
efficiencies are 92.5 and 91.0% for two one-hour tests, respectively.

5.2  PAPER COATER, PLANT D

     Plant D is a typical example of a facility with two coating lines
which each coat more than one coating and which are controlled by a single
carbon adsorption system.  In one of the two coating lines, a vinyl plastic
film is first given one anchor coat and then a final coat of pressure-
sensitive adhesive.  The final product is electrical tape.  The second line
                                   34

-------
                                  TABLE 5-3.  GAS-PHASE TEST RESULTS FOR PLANT  C
Sample
Location
Afterburner

Afterburner

Test
No.
Inlet 1
2
Outlet 1
2
Gas Flowrate
(scfm)
17,920

12,500

Gas Temperature
(°F)
342

1250(870)°

VOC Concentration
(vppm as GI , dry)
14,861a
14,088
1,151
721d
VOC Flowrate
(Ib C!/min)
7.888b
7.538
0.433
0.271
        ^Average of two samples - 15,290 and 14,431.

        DDoes not include carbon in the recycled gas  to the oven.

        "Outlet of primary exchanger.
        j
         Average of two samples - 597  and 844.
Co
Ul

-------
first coats an adhesive which laminates nylon strands to a Mylar® film and
then the film is coated with one release coat and one coat of pressure-
sensitive adhesive.  The final product is strapping tape.  A schematic of
the coating lines is shown in Figure 5-2.  Plant operating parameters are
given in Table 5-4 and characteristics of the coatings are in Table 5-5.

     The Plant D facility was designed to ensure maximum capture of solvent
VOC emissions.  The coater room is maintained at a slight negative, pressure
with respect to the ambient outside pressure.  All of the drying/curing
ovens are either ducted directly to the main header to the carbon adsorption
system or they are ducted to another oven which is subsequently ducted to
the main header.  There are also several hoods, primarily over the coating
applicators, which are ducted to the main header to the carbon adsorber
system.  The overall effect should be essentially 100 percent VOC capture.

     The carbon adsorption system is a typical three-bed system which was
installed in 1974.  A schematic of the system is shown in Figure 5-3.  It
operates with two beds adsorbing and one bed regenerating.  The beds are
switched on-and-off by VOC monitors in the exhaust stacks.  When the VOC
breakthrough concentration is reached, i.e., about 40 ppm toluene, the bed
is shut off and steamed.

     The only gas-phase analyses performed at Plant D were measurements of
the gas flow rate and VOC concentrations in the carbon adsorber exhaust
stacks.  Each bed had an exhaust stack.  The gas flow rates through each bed
were nearly identical (about 15,000 scfm); therefore, all the gas measure-
ments were averaged for all three stacks.  The results are shown in Table 5-6.
                                    36

-------
Line #1
Line
                                              Main  Header to Carbon Adsorbers
        Unwind
                                              Main Header to Carbon Adsorbers
                                                                                                  Steam
                                                                                                  Heaters
                                                                                                  Fans
-fi?

-Jer, c
-"il
i^- , 	 Oven 	 Q
^ Ix^'l - '
Hood

j

>r '•*
"xT'-'n Ct^


                                               fi-
                 Mylar
                 Unwind Vv Nylon Unwind
,	J	[
   - Oven-
                                                                    Hood J
    "»Laminate
-ci   Adhesive  Coater
                                              Release
                                              Coater
                                        - -3      P-j	Oven	
                                        -   I     / *
                                            L    / ^
                                            b	a
                      Pressure Sensitive
                      Adhesive Coater
                                                                                                Wind
                Figure 5-2.    Schematic of  Plant D  coating  facilities.

-------
                  TABLE 5-4.  PLANT D OPERATING PARAMETERS
Coating Line Operating Parameters




                                        Line #1	Line #2




     Web substrate                       PVC                 Mylar




     Line speed (fpm)                     45                 55.5




     Line width (inches)                  36                 59




     Coating Characteristics              (see Table 5-5)






Carbon Adsorber Operating Parameters




     Number of beds - 3 (2 on/1 off)




     Weight of carbon per bed (Ibs) -  7,000




     Superficial velocity (fpm) - <100




     Steam Consumption (Ib steam/lb recovered solvent) -  Not measured
                                     38

-------
                                   TABLE 5-5.   PLANT D COATINGS CHARACTERISTICS
CO
MD

Type of
Line // Coating
1 Anchor
Coat
1 Pressure
Sensitive
Adhesive
2 Laminate
Adhesive
2 Release
2 Pressure
Sensitive
Adhesive


Sample Location
Adsorber Outlet

Coating weight
(gallons/1000 yd2)
7

45
20

3
40

TABLE 5-6.
Gas Flowrate
(scfm) (wet)
14,940a

Coating density
(gm/cm3)
0.8666

0.9231
0.889

0.8605
0.8761

Pound
Volatile per
Content, wt %
91.0

of carbon
pound of
solvent
0.8986

57.1 0.8873
Total Line
28.1

99.3
27.6
Total
0.8858

0.8685
0.8703
Total Line
Lines #1 &
Solvent
application
rate (Ib Ci/mii
0.621

2.634
//I 3.256
1.133

0.564
2.130
#2 3.826
//2 7.0816
GAS PHASE TEST RESULTS FOR PLANT D
Gas Temperature
(°F)
72
VOC Concentration VOC Flowrate
(ppm as C, dry) (Ib C/min)
1,692 c
1
.54 b
         Average gas flowrate per exhaust stack.
         Total for the two adsorber exhaust stacks on line.


        CEaacd on average of 2018,  2029,  2203,  1214,  919  and 1767.

-------
                                                                   Adsorber Exhaust Stacks
                                                 Adsorber Outlet
                                                 Sample Points
                                                                                                       Recovered  Solvent
-P-
o
                       Solvent Laden
                       Air from   	
                       Lines 1 and 2
                                                                                                         Wastewater
                                    Filter/Cooler
                                     Figure  5-3.   Schematic  of Plant  D carbon adsorber system.

-------
      Plant D could not  estimate their  overall  percent  recovery  of  the
 solvents used in the coatings  as originally expected.   Therefore,  the con-
 trol device performance was  calculated assuming  100%  capture of the 7.1  Ibs/
 min of carbon fed to the two coating lines.  The 1.5  Ibs/min of carbon
 leaving in the carbon adsorber exhaust streams indicates  an  efficiency of
 only 79%.   There is no  obvious explanation  for this low value but  it  is
 supported by current above-average  solvent  purchases by the  plant.  The
 analysis of six exhaust samples averaged  1692  ppm C or equivalent  to  242 ppm
 toluene.   This is significantly higher than the  104 ppm toluene level set
 on the hydrocarbon breakthrough detector.   If  the detector were out of
 calibration or otherwise not receiving a  representative sample,  it  would
 permit operation beyond the  desired range before switching the  beds to the
 steaming cycle.   Recent routine testing of  a carbon sample from the bed  by
 the equipment manufacturer did not  indicate any  deterioration of the  carbon.
 The plant is investigating the possibility  of  a  partial by-pass  of  gas flow
 around the carbon beds.

 5.3  PAPER COATER,  PLANT E

      Plant  E  is  an  example of  a well-designed  system for VOC  capture.  Plant
 E  consists  of two  paper  coating  lines which  are  ducted  to a single  carbon
 adsorption  system.   The  coating  lines are relatively small and produce a ni-
 trocellulose-coated  paper which  is used as stencil paper for mimeograph ma-
 chines.  A  schematic of  Plant  E is shown in Figure 5-4.  The normal plant
 operating parameters are shown in Table 5-7.

     As with Plant D, Plant E operates  with their coating  room at a slight
negative pressure with respect to the outside ambient  pressure.   The overall
effect is for all of the oven makeup air to come  from the  coater room and be
pulled through the drying oven.  This  should effect 100 percent  VOC capture
of the solvent used in the coatings.
                                    41

-------
             Solvent Laden Exhaust
                 Oven
Coating Line 111

-------
                 TABLE 5-7.  PLANT E OPERATING PARAMETERS
Coating Line Operating Parameters

Web substrate
Line speed (fpm)
Line width (inches)
Coating solids material
Coating weight (lb/3000 ft2)
Coating density (gm/cm )
Solvent content (wt %)
Pound carbon per pound solvent
Solvent application rate

Line #1
paper
60
34
nitrocellulose
30
0.8894
71.1
0.748
0.9045

Line #2
paper
60
34
nitrocellulose
32
0.8858
66.6
0.748
0.9035
       (Ib G! /min)

Carbon Adsorber Operating Parameters
     Number of beds - 3 (2 on/1 off)
     Weight of carbon per bed (Ibs) - 6000
     Superficial velocity (fpm) - 90
     Steam consumption (Ib steam/lb recovered solvent) - 5-6
                                      43

-------
     The adsorber system has three beds with a single exhaust stack.  The
beds operate with two adsorbing and one regenerating and their switching is
based on VOC breakthrough.  During the test periods, a typical bed would
adsorb for one hour and steam for one-half hour.   Gas-phase analyses were
only performed at the carbon bed exhausts.

     The results of the gas-phase analyses are shown in Table 5-8.  The VOC
concentration represents the average of six one-hour tests using EPA Method
25 over a five-and-one-half-hour period.  Both coating facilities operated
for nearly the entire period.  No sampling problems were encountered during
the test period.

     The plant engineering staff routinely monitors the performance of
the solvent recovery, distillation, storage, and feed systems.  They also
determine the overall solvent recovery from flow measurements.  For the 12
months ending November 30, 1980, the average recovery was 92.14 percent.
Although this value varies slightly from month to month, due to the effect
of temperature on the efficiency of the carbon adsorber, it has been noted
that for the months of April, May, September, and October, the recovery
approximates the annual average.

     Based on the calculated carbon mass rate to  the coaters of 1.808 Ib/min,
the adsorber carbon exhaust rate of 0.032, and a  92 percent overall recovery,
the carbon rate to the adsorber was calculated at 1.696 Ib/min.  This indi-
cates a 98 percent recovery across the carbon adsorber and a 93.8 percent
capture efficiency.   The 6% not captured may be accounted for as residual
solvent in the product leaving the ovens,  solvent handling losses, and
fugitive losses from the coating line.

     The carbon beds and the VOC breakthrough monitors are therefore
very effective, even though the carbon has been in service for six years.
The plant routinely measures the length of time the beds are on stream
before breakthrough to follow the gradual  loss in adsorption capacity
caused by the buildup of heavy organic or  other coating components that
are not readily desorbed in the steam cycle.
                                    44

-------
                           TABLE 5-8.  GAS-PHASE TEST RESULTS FOR PLANT E
                        Gas Flowrate     Gas Temperature     VOC Concentration     VOC Flowrate
    Sample Location       (scfm)              (°F)             (ppm C, dry)          (Ib C/min)


   Adsorber Outlet        10,350               90                   104*             0.0328
^Average of six test runs: 159, 120, 58, 153, 90, and 42 ppm carbon.

-------
5.4  CAN COATER, PLANT L

     This can coating facility applies a single enamel coat to sheets of
tin plate which are used to make three-piece cans.   After being coated with
the solvent-based enamel,  the tin plate sheets  are  dried in a wicket-type
oven.  As shown in Figure 5-5, the exhaust from the drying oven is ducted
to an afterburner.  At the end of the oven are  two  cooling zones where
outside room air is drawn in over the hot sheets to cool them before stack-
out.  Both of the cooling zones are exhausted directly to the atmosphere.
Additional information on the coating facility  and  afterburner is given
in Table 5-9.

     Gas-phase samples were taken at the afterburner inlet, afterburner
outlet, and both cooling zones.  A velocity traverse could not be done at
the afterburner inlet because there was no suitable site.  However,  the
afterburner did not use any additional air for  combusting the solvents in
the oven exhaust gases.  Therefore, the afterburner exhaust rate was used
to estimate the afterburner inlet flow rate. The sample ports at the
afterburner exhaust and cooling zones were satisfactory for both velocity
and VOC measurements.  The results of the gas-phase analyses are shown
in Table 5-10.

     When the amount of solvent ducted to the afterburner inlet plus that
in the cooling zone gas streams is compared with the amount of solvent
applied at the coating applicator, the VOC capture  is calculated at 221
percent.  Obviously, this number is in error.  The  error was initially
thought to be in the data on the applied coating.  However, all of these
numbers were verified by Plant L as correct. No explanation for the poor
material balance has yet been determined.  From the gas-phase analyses,
the afterburner VOC destruction efficiency is calculated at 73 percent.
                                    46

-------
                                                       Cooling Zone A
                                                       Sample Point
                                   Afterburner Outlet
                                   Sample Point
                                    Cooling Zone B
                                    Sample Point
      Roof Line
             Afterburner Inlet
             Sample Point
Uncoated
Sheet Feed
                                   Afterburner with
                                Primary Heat Exchange
                                                _ _.  Coated Sheet
                                                ~S-.i Stackout
                Enamel Coating
                 Applicator
Drying Oven
Cooling
 Zone A
Cooling
 Zone B
           Figure 5-5.    Schematic of Plant  L can coating facility.

-------
                  TABLE 5-9.  PLANT L OPERATING PARAMETERS
Coating Line Operating Parameters


     Sheet material                         -  Tin plate


     Line speed  (sheets/min)                -  105


     Sheet dimensions                       -  33.125" x 33.369"


     Coating material                       -  enamel  (1.0355  gm/cm3)


     Coating weight  (mg/in2, wet)           -  5


     Solvent content, Wt %                  -  22.7


     Solvent composition (Ib C/lb solvent)  -  0.8483


     Solvent application rate  (Ib C/min)    -  0.246



Afterburner Operating Parameters

     Type of afterburner                    -  thermal oxidizer


     Solvent laden air to afterburner  (scfm)-  7,750


     Afterburner temperature (°F)           -  1200


     Residence time  (sec)                   -  1.8*
_

 Estimated by plant  engineer
                                      48

-------
                                   TABLE 5-10.   GAS-PHASE TEST RESULTS FOR PLANT L
-p-
Sample Location
Afterburner inlet
Afterburner outlet
Cooling zone A
Cooling zone B
Gas Flowrate
(scfm)
7,750*
7,980
13,330
10,800
Gas Temperature
(°F)
2200
787
155
118
VOC Concentration
(ppm as C, dry)
1447
389
220
336
VOC Flowrate
(Ib C/min)
0.343
0.0922
0.0897
0.111
           Estimated from incinerator exhaust flowrate.

-------
5.5   CAN COATER,  PLANT M

     Two separate and different coating lines were tested at Plant M.
The lines are designated as Line #1 and Line #2 and are discussed in the
following subsections.

5.5.1  Line //I

     Coating Line #1 consisted of two litho printers and one lacquer
coater (Figure 5-6).  The litho printers are fed base-coated, enameled
sheets and use non-solvent-type colored inks to print the designs or
wording of the can label.   The sheet may be run through the litho presses
one or more times depending on the desired number of colors; however, on
the final print,  the label is sealed with a solvent-based lacquer coat.
This lacquer coat will protect the finish and give the can a glossy
appearance.

     The solvent emissions from the drying oven were ducted to a thermal
afterburner.  This afterburner operated at relatively low temperatures
(500 to 600°F) and had no  primary heat exchange.   At the end of the oven
was a single cooling zone which was ducted directly to the atmosphere.
This unit was designed for smoke and odor control rather than VOC
emission reduction.

     During the testing,  a lacquer coat was being applied.  Table 5-11
gives the normal plant operating parameters during the test period.

     Gas-phase analyses were performed at the afterburner inlet, after-
burner outlet, and cooling zone exhaust.  The results of these tests are
given in Table 5-12.  It was difficult to make good velocity measurements
at any of the test sites.   All of the test locations, except at the
afterburner inlet, had duct runs which were shorter than those recommended
in EPA Method 1.  Therefore, rather than the normal 24 traverse points.
                                   50

-------
  Outlet Sample
      Point
                      Roofline
                      Inlet Sample
                          Point
                                    Exhausted to
                                    Atmosphere

Sheet
Feed

8> & cP ^
U O o --'7
Lttho Lacquer /




Presses Coater

Final Control
Sheet
               Drying Oven Cooling Zone
Figure 5-6.   Coating line 1,  Plant M.

-------
            TABLE 5-11.  PLANT M OPERATING PARAMETERS - LINE #1
Coating Line Operating Parameters




     Sheet material




     Line speed (sheets/min)




     Sheet dimensions




     Coating material




     Coating weight  (mg/4 in2)




     Solvent content, Wt %




     Solvent composition (Ib C/lb solvent)   0.7751




     Solvent application rate (Ib C/min)     0.227




Afterburner Operating Parameters




     Type of afterburner                     Thermal




     Solvent laden air to afterburner  (scfm) 4,400




     Afterburner temperature (°F)            550




     Residence time  (sec)                    Unknown
Tin plate




68




33" x 34"




Lacquer (0.935 gm/cm3)




12




58.0
                                     52

-------
Ul
Co
                               TABLE 5-12.  GAS-PHASE TEST RESULTS FOR PLANT M -  LINE  #1
Sample Location
Afterburner inlet
Afterburner outlet
Cooling zone
Gas Flowrate
(scfm)
4,400
3,170
12,500
Gas Temperature
(°F)
148
551
102
VOC Concentration
(ppm as C, dry)
2,616
2,678
125
VOC Flowrate
(lb C/min)
0.349
0.260
0.0478

-------
the velocity was measured at 48 points.  Even with this added data, the
velocity measurements may be suspect.  This is indicated in the data (see
Table 5-12) which shows a much lower gas flow rate leaving the afterburner
than the gas flow rate entering.

     Another operational problem which hindered the gas-phase test work was
the frequent stoppages of the coating line.  During the test date, several
mechanical problems were encountered with Line #1"; therefore, the coater was
turned on and off several times.   The Century  OVA was used to monitor fluc-
tuations in VOC concentrations during the test run.  The strip chart,  Figure
5-7, shows typical fluctuations experienced at Line #1.

     Due to the questionable velocity readings and the many facility
startups and shutdowns, the material balances for the test period were very
poor.  The percent VOC capture was calculated at 174 percent and the VOC
destruction across the afterburner was calculated as 26%.  However, if the
flow rates were assumed to be equal  (since no additional air is added at the
incinerator burner) and only the VOC concentrations are considered, i.e.
2616 vs. 2678 ppm, the VOC destruction is essentially zero.  This may be true
for an afterburner operating at such a low temperature  (551°F).

5.5.2  Line #2

     Coating Line #2 consists of a single coater and a single oven  (Figure
5-8).  The oven consists of six drying zones and one cooling zone.  The
gases in the drying zones flow countercurrent to the coated sheets.  Finally,
in Zone 1, the oven gases are exhausted to a catalytic oxidizer.  Table 5-13
gives the operating parameters for Line #2 during the test date.
                                     54

-------
Ol

Ul
03
V—'





O
       If ,000 --•-
        1,000
          100
           10 -—
            1 ._
                                    T-,— •---:- j~' .-•--. -
                                          15
                                                               30


                                                       Time  (minutes)
                                                                                     r   -
45
-4-
 60
                    Figure 5-7 ,  Incinerator inlet VOC concentration  (as CHO versus  time,  Plant M,  Line #1

-------
           o
          Coater
Sheet Feed
                                                                      \  I  Outlet
                                                                       \J Sample
                                                                           Point
                                                                         \
                                                                           xRoofline
                                                                  Catalyst Bed
                        Zone
   v  Exhausted  to
     Atmosphere
Cooling
 Zone
                Final
             Control Sheet
                                   Figure  5-8.   Coating line 2,  Plant M.

-------
             TABLE 5-13. PLANT M OPERATING PARAMETERS - LINE #2
Coating Line Operating Parameters




     Sheet material                          Tin plate




     Line speed  (sheets/min)                 100




     Sheet dimensions                        33" x 34"




     Coating material                        Enamel (1.3744 gm/cm3)




     Coating weight  (mg/4 in2)               50




     Weight percent solvent                  34.1




     Solvent composition (Ib Ci/lb solvent)  0.7248




     Solvent application rate (Ib Cj/min)    0.764






Afterburner Operating Parameters




     Type of afterburner                     Catalytic




     Solvent laden air to afterburner  (scfm) 6,930




     Afterburner temperature (°F)            550




     Residence time (sec)                    Unknown
                                      57

-------
     Gas-phase samples were taken at the afterburner inlet, afterburner
 exhaust, and codling zone exhaust.  The test results are shown in Table 5-14.
 As with Line #1, the velocity sample sites were not suited for accurate
 measurements; however, they were better than those on Line #1.  Line #2
 did  experience process fluctuations but not as many as Line #1.  The
 Century OVA hydrocarbon monitoring data for a typical operating hour for
 Line #2 is shown in Figure 5-9.

     The material balance for Line #2 shows a potential 90% VOC capture
 efficiency.  This includes 74% VOC captured in the oven which is routed to
 the  afterburner plus 16% VOC which is currently exhausted from the cooling
 zone.  The VOC reduction across the catalytic afterburner is calculated at
 49%.  This low value may be attributed to the relatively low operating
 temperature of 550°F rather than 600°F+ normally used for catalytic after-
 burners.  This operating temperature was previously set to control odors
 rather than for VOC destruction according to the plant management.  The
 potential overall efficiency is 44% (90 x 49%).

 5.6  CAN COATERS, PLANTS OUTSIDE OF ILLINOIS

     Because of limited time and resources,  field tests were conducted
 at only two sites in Illinois.  In order to provide more data, studies
 performed at seven other facilities outside Illinois by other companies
were also considered.   The results of these tests are presented in the next
 section, in Table 6-1,  and are used to support  the conclusions and
 recommendations.   Five studies were performed using EPA Method 25, one with
 a flame ionization detector (FID)  and one by means of a solvent material
balance.
                                    58

-------
TABLE 5-14.   GAS-PHASE TEST RESULTS FOR PLANT M - LINE #2
Sample Location
Afterburner inlet
Afterburner outlet
Cooling zone
Gas Flowrate
(scfm)
6,930
6,350
27,310
Gas Temperature
(°F)
228
548
94
VOC Concentration
(vppm as GI , dry)
2,696
1,517
144
VOC Flowrate
(Ib Ci/min)
0.566
0.291
0.119

-------
PC)
u
s
     10,000
1,000—
        100
         10-


                                                     -J-

                                                                                           (
                                                                               	i : .11 i! f t j jij.^ijj-^i	
                0                       15                       30                      45                        60
                                                         Time (minutes)
                Figure 5-9,   Incinerator  inlet VOC concentration  (as  CHit)  versus time, Plant M, Line #2

-------
                                 SECTION 6
                      CONCLUSIONS AND RECOMMENDATIONS

     Because of the small number of surface coating lines that  were studied
in this program, as well as the limited number of samples taken,  it is  diff-
cult to draw conclusions from the performance of these specific lines and
say that they are fully representative of any other facility or even the
average of the others in Illinois.  The best that can be expected is that
the results may actually be representative of the sites tested.

     However, from the observations made during the course of this program
it is possible to make some judgments as to the performance that  might
be expected from surface coating facilities if reasonable collection and
abatement equipment were provided and proper operational and maintenance
procedures were followed.

     The primary objective of the testing was to measure the VOC collection
efficiency and the control device efficiency at each coating line.  While
this was accomplished, there were several notable exceptions that are
discussed in the following paragraphs.  The fact that the performance
observed at some facilities was significantly below that proposed by either
the EPA CTG guidelines or the Illinois SIP Rule 205(n)(2)(A), does not
mean that better performance is not possible, and this will also be
discussed.

     A by-product of this study was the opportunity to evaluate the metho-
dology for determining the performance of a surface coating facility.  In
particular, the recently developed EPA Method 25 to measure the
total gas nonmethane organic carbon concentration in gaseous streams was
                                    61

-------
simple and cost effective.   The field sampling procedures were straight-
forward and easily performed by those experienced in' routine sampling of
gaseous streams.   The analyses were performed in off-site laboratories
using established dry gas instrumental techniques such as gas chromatographic
separation and oxidation and reduction.   The potential problems of sample
deterioration or diffusion through gas bags were precluded by Method 25.
Alternate sampling and analytical methods would have been more costly,
especially if speciation of the sample had been performed to achieve the
necessary accuracy.

6.1  PAPER COATING OPERATIONS

     From a review of the test data developed at three sites in this program,
C, D, and E, plus recent data obtained by other investigators at sites A,
B, F, G, H, and I, as presented in Table 6-1, it is possible to draw some
conclusions about collection and control efficiencies.

     Collection Efficiency

     At sites C and E, the collection of VOC emissions was determined to
be 100 and 94%, respectively.  At site B, where the overall efficiency was
determined to be 90%, if it is assumed that the carbon adsorber had a rela-
tively high efficiency of 98%, then the capture must have been at least 92%.
Unfortunately, the test results at D, A, and F do not permit a determination
of the capture efficiency.   However,  if the coating rooms are operated under
a slight negative pressure as was observed in Plants C, D, and E, and if
the coating applicators are close coupled to the drying ovens, then the
capture efficiency of paper coaters can normally be expected to be at
least 90%.
                                   62

-------
                       TABLE 6-1.  SUMMARY OF RECENT TEST DATA AT SURFACE COATING OPERATIONS
          Plant 10
 Type of
VOC Control
VOC Test
 Method
VOC Capture and abatement efficiencies
% Capture^     % Controlb    % Overallc
ON
US
       Paper Coating
         Facilities

             A*                  TO
             B*                  CA
             C                   TO
             D                   CA
             E                   CA
             F*                  TO
             G                   CA
             H                   TO
             I                   TO

       Can Coating Facilities
                Flame ionization
                Solvent balance
                Method 25
                Method 25
                Method 25
                Method 25
                Material balance
                Flame ionization
                Flame ionization
                      NA
                      NA
                     100
                      NA
                      94
                      NA
                      NA
                      NA
                      NA
                 96-97
                  NA
                  96
                  76
                  98
                  85
                  NA
                 97-99
                 98-99
  NA
  90
  96
  NA
  92
  NA
93-96
  NA
  NA
L
M (Line #1)T
M (Line #2)
PH*
Q *
R *
se*
TO
TO
CO
TO
-
TO
TO
Method 25
Method 25
Method 25
Method 25
Method 25
Method 25
Method 25
221
174
90
77
74-79
73
12
73
26
49
NA
NA
89
NA
NA
NA
64
NA
NA
65
NA
       TO = Thermal oxidizer, CO = catalytic oxidizer, CA= carbon adsorber, NA = not available.

        % capture is based on the sum of all solvent which leaves the facility in ducted gas streams.  For
        can coating operations this includes the cooling zones and enclosed conveyers or elevators.

        % control is the VOC reduction across the control device.

       /-•
        % overall is the maximum potential reduction of VOC, i.e., capture x control.

        Base coat applicator for a three piece can.

       Q
        Inside spray operations.

       *Test programs conducted earlier by other investigators.

-------
     Control Device Efficiency

     The five thermal oxidizers shown in Table 6-1 were determined to
operate at 96, 96,  85, 97,  and 98% efficiency at Plants A,  C,  F,  H,  and I.
The design parameters of afterburners have been well developed and if
properly operated at 1400°F with 0.3 seconds  or more residence time, these
units can routinely perform at 90-95% VOC destruction efficiency.   The
lower efficiency at Plant F might be attributed to an operating temperature
(1200°F) below the design temperature.

     The carbon adsorbers at Plant E were determined to be 98% efficient.
At Plant B, if the capture efficiency was 99% or better, the carbon adsorber
would have been operating at about 91% to have achieved the reported 90%
overall control efficiency.  This seems reasonable since carbon adsorbers
have routinely been shown to be operated at 90-98% efficiency.

     Carbon adsorbers lose capacity with age due to the buildup of heavy
ends that are difficult to desorb during the steam regenration step.   How-
ever, if a detector is used to signal the VOC breakthrough, it is possible
to operate with a reasonably high efficiency, i.e., 90+%.   When the on-
stream cycle time becomes short, and the steam consumption becomes excessive,
e.g., 8 pounds/pound of solvent recovered, then it is time to  change out
the carbon in the beds.  At Plant D, the relatively low efficiency,  may
have been due either to a general loss in bed capacity, operating beyond
the breakthrough point, insufficient steam during desorption,  or the VOC
detector being out of calibration.  If operating problems are resolved,
90% or better recovery may be achieved.

     Overall Control Efficiency

     In at least four cases, the measured overall control was 90% or better.
Based on an expected minimum capture efficiency of 90% as discussed above
and a control device efficiency of at least 90%, the overall efficiency of
a properly designed and operated paper coating line would be 81% or higher
(90% x 90% = 81%).

                                     64

-------
     Fugitive Emissions

     Instantaneous measurements of hydrocarbon concentration were taken
in a survey of the production areas using a portable organic vapor analyzer
(OVA).  Figure 6-1 depicts the typical concentration of solvent vapor in
the ambient air of a paper coating room.  As would be expected, the con-
centration was highest directly above the coating applicator, and not as
high in the flash-off area.  This indicates that the solvent vapors tend
to be carried along with the web and are drawn into the drying oven if it
is operated with an adequate draft.  It should be noted that even after
drying, the final product continues to emit some small amount of solvent
vapor.  Earlier studies have indicated that generally 1  to 5% of the
solvent used at the coater may still be present in the product leaving
the oven.

6.2  CAN COATING OPERATIONS

     At the three tested can coating facilities, the only surface coating
operation that used an add-on control device was sheet coating.  This is
apparently because it:

        is a major step in fabricating three-piece cans;
        is a continuous rather than intermittent operation;
        uses a high solvent rate at the applicator; and
        uses an oven to dry or bake the finish from which its
        exhaust can be easily ducted to a control device.

     Other coating operations which do not generally have add-on
controls, although they may have hoods exhuasting directly to the atmosphere,
are those which:
                                    65

-------
   0 ppm
 Substrate
  Feeder
                            40-80 ppm
      100-300 ppm        @ 12 inches above
@ 12 inches above center    substrate
 Coating*
Applicator
                             Flash-off
                               Area
100-800 ppm
  in duct
   Drying/
   Curing
    Oven
                                                                20-40 ppm
 Final
Product
Notes:  Readings were measured by a Century OVA meter, which was calibrated with
        methane.

         Readings directly over the applicator rolls and drip pan ranged from 800 to 1500 vppm
         depending on volatility of solvent.
Figure 6-1.  Typical measurements of VOC concentration in vicinity of can and paper
             coating lines.

-------
     •  are performed only intermittently, i.e. are not used for all
        product lines,

        use a low solvent application rate,

        use a waterborne coating or one requiring a UV cure, or

     •  do not require an oven.


Examples of these uncontrolled operations are:


     •  two- and three-piece can interior body spray,

     •  two-piece can exterior end (spray or roll coat),

     •  three-piece can side-seam spray, and

     •  end-sealing.


     Collection Efficiency


     The VOC capture or collection efficiency at a can coating line is

more difficult to determine accurately than that of a paper coater for

several reasons:


     •  the operation is subject to more frequent interruptions,
        especially if it is an old facility,

     •  the coating formulations are more complex,

        some coatings are relatively thin and hence the solvent
        rates are low, and

        oven cooling zones exhaust large volumes of air that may
        contain VOC vapors.  This is possible if solvent vapors
        pass from the drying zone to the cooling zone because of
        an imbalance in pressure between the drying and cooling
        zones.  Also, VOC's will enter the cooling zone if there
        is inadequate drying,  resulting from too low an oven
        temperature, too short a residence time, or a heavy coating.
                                   67

-------
     For these reasons, the observed efficiencies ranged from 12 to over
200%, but four were between 73 and 90%.   The actual capture efficiencies
of can coating operations are apparently lower than that observed for paper
coating partially for the reasons cited above.  Also,  the facilities tested
were not originally designed to recover fugitive VOC emissions.   There are
generally no hoods or floor sweeps positioned near the applicators and
flash-off zones.   Where solvents with a high molecular weight are used,  the
presence of floor sweeps can enhance the capture efficiency.

     Add-On Control Device Efficiency

     Because of the wide variety of solvents used in can coating formula-
tions,  the fact that several solvents may be used at one time and the fact
that some are water soluble, carbon adsorbers are not generally  used to
recover VOC emissions.   Rather, thermal or catalytic afterburners are used
to destroy the emissions.
     The efficiencies of the afterburners tested in this  program were not
typical of well designed and operated afterburners.   In all  three instances,
the operating temperatures were significantly below that  required to  fully
combust hydrocarbons.  Thermal oxidizers should operate at 1400-1500°F with
a residence time of at least 0.3 seconds to be effective.  The design
operating temperature for catalytic oxidizers is a function  of the particular
catalyst used, but is generally greater than 600°F.   The  catalyst in  this
type of unit is subject to poisoning and should be replaced  when the  per-
formance deteriorates to such an extent that increasing the  operating
temperature is ineffective in maintaining a high VOC destruction efficiency.

     The 89% efficiency of the thermal unit at site R is  close to the
90 to 95% values achievable by properly functioning systems.
                                   68

-------
     Overall Control Efficiency

     None of the can coating facilities tested in the program have an
overall control efficiency in excess of 65%.  Since it is feasible to
operate afterburners at 90-95% efficiency, it is the design and operation
of the VOC capture system that is the more critical step in achieving a
75 or 81% overall control efficiency.  The emission sources are generally
categorized as being either from the coater/flash-off areas or from the
curing ovens.  The proposed NSPS for the beverage can surface coating
industry13 indicates in §60.493 the distribution of emissions between
coater/flashoff and curing oven areas shown in Table 6-2.  The values are
based on information presented in the can coating CTG1"*.   It further states
on page 78989 of Reference 13 that industry representatives agreed that the
values are representative of the industry.

     The ease of controlling these emissions is site-specific, and
although the oven emissions are simpler to control since they are already
contained in a vessel,  there are several design and operating parameters
to keep in mind to ensure that the capture of the emissions from any source
are maximized.   Modifications that  might enhance the capture of VOC
emissions in curing ovens are:
        Seal the coating room so that it is always under a slight
        negative pressure.   This assures that all fugitive VOC
        emissions are either captured by hoods or floor sweeps or
        drawn into the baking oven.
        Maintain the oven under the  lowest absolute pressure so that
        the vaporized VOC's do not escape into the coating room or
        into any subsequent cooling  zone with a separate exhaust
        into the atmosphere.
        Minimize inlet and outlet openings, inspection ports,  etc.,
        on the oven to reduce air flow requirements,  yet remain below
        the recommended 25-40% LEL limits set for safety reasons.
                                    69

-------
                TABLE 6-2.  DISTRIBUTION OF VOC EMISSIONS*
                                             Emission Distribution
                                          Coater/              Curing
	Coating  Operation	flashoff	Oven


 2-piece aluminum  or  steel cans:
   Exterior  base coat operation               0.75               0.25
   Overvarnish  coating operation                .75                 .25
   Inside spray coating operation               .80                 .20

 3-piece steel  cans:

   Exterior  base coat operation                 .10                 .90
   Interior  base coat operation                 .10                 .90
   Overvarnish  coating operation                .10                 .90
   Inside spray coating operation               .80                 .20

 Steel ends:
   Exterior  coating operation                   .10                 .90
   Interior  coating operation                   .10                 .90


 ^Reference  13
                                     70

-------
     The collection of VOC from the coating/flash-off areas requires
more ingenuity because the operations frequently are of necessity
spread out.  Until recent years, efforts to consolidate these operations
and capture emissions were minimal.

     Where it is not feasible to seal the entire coating room, collection
of emissions from the coater flash-off area may be enhanced by using hoods
mounted above the coater and/or conveying line, slot hoods closely mounted
adjacent to, and at the same elevation as, the coating line, or "floor
sweeps" which are collection devices mounted at floor level to capture
heavy VOC vapors.  Also, the installation of baffles will minimize cross
drafts in the flash-off area.

     The exhaust from the drying or curing zone of the oven is the
richest in VOC,  as demonstrated at Plants L and M, and is most suitable for
control in an afterburner.  The exhaust from the cooling zones may also
contain VOC vapors, but are in lower concentration and frequently large in
volume relative to the drying zone exhaust.

     Fugitive Emissions

     Instantaneous measurements of hydrocarbon concentration taken at
the two plants reflected a pattern similar to that noted at paper coaters
and shown in Figure 6-1.  It indicates that some of the solvent vapors
flashed-off ahead of the oven may be drawn into the oven if there is an
adequate draft.   Although there is no solvent adsorbed in the metal sheets,
the surface coating may account for the release of a slight amount of
residual solvent vapors.  It is the application and flash-off zones
that deserve the greatest attention in the design of hoods for capture of
vapors and location of baffles to minimize entrainment of fugitives by air
currents.
                                    71

-------
 6.3  ANALYTICAL PROCEDURES FOR EVALUATING EMISSION CONTROL SYSTEMS

     Ideally, emissions of volatile organic compounds would be expressed as
 the sum of the masses of the individual components.  However, in practice,
 many emission sources contain a variety of chemical species and it is
 difficult to identify and quantify them without expending considerable time
 and effort on the laboratory analyses of gas or liquid samples.  Simple
 techniques using a GC with an FID can be used to estimate the concentration
 of several components, collectively.  The concentration can be expressed
 as equivalent methane, ethane, or propane.  This procedure is subject to
 inaccuracies because of the wide variation in detector response factors
 for individual components.

     Therefore, a more direct approach that is being more widely accepted
 is to express the emissions as the mass of carbon emitted.  It uses the
 recently developed EPA Method 25 for "Total Gaseous Nonmethane Organics"
 (TGNMO).  This procedure was first used in California in the early 1970's
 and has been well accepted.   The field sampling procedure is relatively
 simple and can be easily learned.   The analytical procedures are relatively
 fast and the results are reproducible, but they require a moderately complex
 and expensive gas chromatographic apparatus with oxidation and reduction
 capabilities.  As Method 25 gains more acceptance nationwide, the number of
 laboratories providing this service will probably increase and the costs
may decrease.  There have been several studies initiated to further
 evaluate Method 25.

     EPA Method 25 is suitable for analyzing gas samples such as the
vapor collected in a hood leading to a control device and the exhaust from
 that device.   This permits a ready and accurate calculation of control
 equipment efficiency.  Unfortunately,  as yet,  there is no accepted chemical
method for analyzing a liquid coating  mixture to determine the carbon con-
 tent of the volatile fraction.   The value can be calculated from a knowledge
 of the coating formulation;  i.e.,  its  volatile content by weight,  and the
                                    72

-------
 weight or volumes of individual solvents used in the mixture.   However,
 this detailed information is not always available especially  if a  "nominal"
 solvent blend is recycled to the mix vessel  and components  are  added "on
 the run" to bring the physical performance of the coating into  the desired
 ranges.  It therefore appears necessary to develop an acceptable laboratory
 procedure for determining the carbon content of coating  liquids so that
 the calculation of collection efficiency and overall efficiency may be per-
 formed on a consistent basis, i.e.,  a carbon mass balance.

      In this  test program, one manufacturer,  Plant  C,  provided  a calculated
 value for the carbon content,  but for  the  other  sites, an attempt was made
 to  obtain the carbon content  of the  coating  mixture by using an oxidation-
 reduction technique  (as in Method 25)  on the solvent fraction distilled
 from the mixture.  Although  the recovery of  solvent  in the distillation
 ranged from 38  to  98%,  the carbon content  by analysis  agreed closely with
 that obtained from calculation of the  nominal formulation provided by the
 plants as indicated  in  Table  4-1.

     Using  the mass of VOC as  the basis  for calculating performance has
been  the  recommended procedure  in many regulations, but if a change in
composition occurs between points of analysis, the calculated values are
less meaningful.  This can be demonstrated by reference to Figure 6-2
where  the capture efficiency, EI, control device efficiency, Ea, and overall
efficiency, £3, are calculated using two different bases, either VOC mass,
or carbon mass.  If the solvent in the coating mixture at Point  A is
initially a 50/50 blend of xylene and butyl cellosolve, the  carbon  con-
centration is 75.8%;  i.e., 5.28 pounds of solvent mix contains 4.0  pounds
of carbon.  However,  if at the oven outlet, B, the ratio  of  solvent components
is found  to be 55/45 reflecting a selective fugitive loss of cellosolve
ahead of  the oven, the carbon concentration is 77.3%.  If it is  further
determined that when these vapors are oxidized in an afterburner, the
remaining exhaust vapors at Point C  contain a significant amount of
                                    73

-------
      0)
      •W
      I
      e
      10
      co
      pi
      o
      td
      o
                         Basis
                     VOC using C
                     VOC using C1
                  Carbon(Method 25)
EQUATIONS FOR CALCULATING EFFICIENCY

       Capture - EI"-T— x 100
                                                                                         Control Device » E2 =  5  x 100
                                                                                                             B
                                                                                          Overall
                                                                                                         x Ea  x 0.01
                                                                                                          x 100
                                                                                                                      100
                                                                            where A « Applicator Rate
                                                                                 B = Carbon Rate to Control Device
                                                                                 C = Carbon Rate from Control Device
                                                                                 Note:  Calculations for adjacent diagram are based on the
                                                                                       following assumed compositions:

Formula
HtX C
Stream
5
E
C
C1
Xyleue
CeHio
90.6

50
55
5
90
        Butyl Cellosolve
            CeHii.02
               61
               50
               A5
               5
               5
Fornulduhyde
  CH20
   40
   90
    5
Solvent
  Mix
Calc'd
 Ht% C
 75.8
 77.3
 43.6
 86.6
                                    Overall
Figure   6^-2.   Graphical comparison of bases  for  calculating emission  control efficiencies.

-------
formaldehyde, the carbon concentration would be only 43.6%.  From these
analyses and a change in VOC mass from 5.28 to 4.66 to 1.83, at A, B, and
C, efficiencies EI, E2,  and E3 based on VOC's are calculated to be 88.3,
60.6, and 53.5%, respectively.  However, the same efficiencies based on
the carbon mass (determined by Method 25) would be 90, 77.8, and 70%.

     As a further example, if the composition of the remaining VOC's
at Point C were 90% xylene, as shown in the table as C1, a significantly
different set of £2 and £3 efficiencies would be reported based on VOC's,
even though the carbon based values would be unchanged.

     These examples show that the calculated control efficiency values
are dependent on an accurate identification and quantification of the
individual VOC components.  Also, since the composition of the mixture of
VOC vapors is most likely to change between the applicator and the after-
burner outlet, the result is an efficiency value which has a mixed or
vague basis of calculation.

     It is therefore recommended that consideration be given to using
carbon concentration,  rather than VOC concentration, as the basis for
calculating the capture, control device and overall efficiencies.
                                    75

-------
76

-------
                                  REFERENCES
 1.  Rolke, R. W., et al.  Afterburner Systems Study.   Shell Development
     Co., Emeryville, California.  EPA Contract No.  ESHD 71-3.
     EPA-R2-72-062, August 1972.

 2.  Industrial Surface Coating  (Can Coating), Emission Test Report.
     Metal Container Corporation, Jacksonville, Florida.  U.S.  EPA,
     EMB Report 79-1SC-8, December 1979.

 3.  Industrial Surface Coating Cans, Emission Test Report.   American
     Can, Forest Park, Georgia.  U. S. EPA, EMB Report 79-1SC-7,
     August 1979.

 4.  Pantalone, J. S. and L. Shepard.  Consideration of Model Rule for
     the Control of Volatile Organic Compound Emissions from Can  and  Coil
     Coating Operations.  California Air Resources Board,  July  26,  1978.

 5.  Nelson, T. P.  Trip Report on Paper Coating Facilities, Manufacturer A.
     Radian Corporation, Austin, Texas (Docket confidential  file).

 6.  Pressure Sensitive Tape and Label Surface Coating Industry,  Background
     Information for Proposed Standards, Draft EIS,  App. C,  pp  3-8.

 7.  Reference 6, App. C, pp 3-7.

 8.  Reference 6, App. C, pp 8-13.

 9.  Reference 6, App. C, pp 3-4.

10.  Reference 6, App. C, pp 3-4.

11.  Federal Register, Vol. 45, No. 36,  February 21,  1980, Rules  and
     Regulations, p 11482.

12.  Package Sorption Systems Study.   MSA Corporation,  Evans City,
     Pennsylvania.  EPA R2-73-202, U.S.  Environmental  Protection  Agency,
     Research Triangle Park, North Carolina,  Contract  EHSD 71-2,  April 1973.

13.  Federal Register, Vol. 45, No. 230,  November 26,  1980.   Proposed
     Rules, Standards of Performance for New  Stationary Sources;  Beverage
     Can Surface Coating Industry.

14.  Control of Volatile Organic EnicsionG fron Existing Stationary Sources,
     Volume II: Surface Coating of Cans,  Coils,  Paper,  Fabrics, Automobiles
     and Light Duty Trucks.  EPA-450/2-77-008,  May 1977.

15.  Reference 6, p  3-14.
                                      77

-------
78

-------
 APPENDIX A
TEST METHODS
    A-l

-------
STANDARD METHOD OF TEST FROM ASTM STANDARDS,  PART 25
             Philadelphia,  Pennsylvania
  1  Standard Method of Test for Density of Paint,  Varnish
     Lacquer, and Related Products
     D-1475-60,  pp.  273-275

  2  Volatile Content of Solvent-Reducible Paints
     D-2369-80,  pp.  D-4-5

  3  Vacuum Distillation of Solvents  from Solvent-Base
     Paints for  Analysis
     D-3272-76,  pp.  611-613
       U.S.  EPA METHODS IN 40  CFR PART  60,
             APPENDIX A,  PP.  88-128
  4  Method 1 - Sample and Velocity  Traverses  for
     Stationary Sources

  5  Method 2 - Determination of  Stack Gas  Velocity and
     Volumetric Flow Rate (Type S Pitot Tube)

  6  Method 3 - Gas  Analysis  for  Carbon Dioxide, Oxygen,
     Excess Air,  and Dry  Molecular Weight

  7  Method 4 - Determination of  Moisture Content  in
     Stack Gases

  8  Method 25:  Determination of  Total Gaseous Non-Methane
     Organic Emissions as Carbon
     Federal Register, Vol. 45, No.  194, October 3,  1980,
     pp.  65959-73.
                        A-2

-------
           APPENDIX B
TEST INSTRUMENT CALIBRATION DATA
   AND DESCRIPTIVE INFORMATION
              B-l

-------
              1.0 4.
              0.8 ..
              0.6
 Magnehelix
   Heading
(inches^of
   water)
0.4
              0.2
Slope = 0.9473

Correlation -0.9999
                                         0.4          0.6         0.8          1.0          1.2
                                    Water Manometer  (inches  of water)
                       Figure A-l.   Calibration  curve  for  one-inch  HaO Magnehelix
                                    (#70923  mm69)
                                                B-2

-------
              5  .,
               4  ..
               3  .,
lagnehelix
  Rending
(inches of
  water)
               2 -,
               1 -•
                                                                   Slope  =  0.9437

                                                                   Correlation = 0.9999
                                                                 1	1	1      i
                                     Water Manometer  (inches of
                      Figure A-2.  Calibration curve  for  five-inch H20 Magnehelix
                                   (#805195524)
                                               B-3

-------
           HCE ENGINEERING
          A •UWtOIARY Of HA^f. INC.
          8826 NORTH LAMAR &LVD.
            AUSTIN. TEXAS 787W
«T8T T«IE
                        / &
£DIY S-Sf
r

RVNKC.
1
2




^,U*V.Stj
*m Hj-0
(in. HjO)
,\
.ur
.fi"

'--^-iAIiC

HMUMUTIG
(in MjO'
.^-f
.3^3
.2of

^^a^


e,w
.83/
,g;£
O /I "7
> O ~ /
.23g
--rz-cszyrr


OtVIATlO*.
, l>llt" ,? 1
-. 05T

• Otfy !

-------
                  SERIES 600Q-P
 ALNOR
INSTRUMENT COMPANY
      B-5
                           Tfitfjpr.', n
                           'FORM.6270^/'

-------
CL
<



§S
Q>Cxi

vS »
   C*


^'T
2sr
o
Q
TJ-
<^
Q
Q
Ci

^
Q
o
\9
  O
* - 
<^
d


S
5
Q
_£
\A
^_n



_
J.
	 ._._... ... _
.X--
s
a
H
ft
i i 	 ^±.
i
^
i

i 1,1
	 1 	 f -• ( I i 1
Q 0 O O
Q O O Qi
^ 0 vS ^
^ -^ — —
^r\
T N^
•^ ^\
XV
\
4_
1

. i
1 * T
Q
Q
CX3





V----1
v\
s
\
( 	 _
O
• v°
cjo


0
. <^
^t

- o.
I p — \ ^^
1 °

    7
    Q

    S
    Ui
    K
    U)

    Q

    si
    Qi
    Q
            PP

-------
--J
                            o
                r
                &
                O
                 r-
                O
                3.   _
                m
                m
                      o
                      o
                      0
                      -^
                      CJ
                      O

-------
POE CO.,INC.
99 REINERMAN^ST. • HOUSTON, TEXAS 77007 « 71M61-3816
                                August 26, 1980
Radian Corporation
P.O. Box 9984
Austin, TX 78766

Ref:  #35711                              -
Sub:  Certification of Calibration -

Dear Sir:          '                 .

This is to certify that the following Rockwell Dry
Gas Meters have been calibrated with  American Bell
Prover #1763, which is traceable to Bureau of
Standards, reference #26727, PI-TAPE.
      #704584
      #715573
      #696761
      #706941
                            *••« »•  /
15 eft.
30 eft.
45 eft.
15
30
45
15
30 ,
45
15
30
45 •
0.0%
6.0%
0.27.
"0.0%
0.0%
0.0%
.97.
.9%
.97.
.2%
0.0%
0.07.
*
4»
+
t
t
t
+
•f
+
.
4
t
 CWP :mv
                                Sincerely,

                                CARL POE CO.,
                                Carl W.  Poe
                          B-8

-------
   APPENDIX C
TYPES OF COATINGS
AND SOLVENTS USED
      C-l

-------
              APPENDIX C.   TYPES OF COATINGS AND  SOLVENTS  USED
Can Coating
Coating
Solvents
     M
Paper Coating

     C

     D
                  Oleo resins (75 wt %)
                  Phenolic
                  Water-based
                  Vinyl
Epoxics
Phenolics
Alkyds
Vinyls
Pigmented inks
Adhesive

Pressure-sensitive
  adhesive

Nitrocellulose
S-Thinner
Cellosolve acetate
Butyl cellosolve
Xylene
Isophorone
H-thinner

X-A (67% xylene, 33% diacetone alcohol)
Lacquer thinner (50% MIBC, 45% xylene)
  5% isophorone
Vinyl solvents (75% xylene, 25% MIBK)
E-Solvent (50% xylene, 50% butyl
  cellosolve)

Confidential
                                           Toluene and others

                                           Toluene, ethyl acetate,
                                           isopropanol
MIBC = methylisobutyl carbinol
MIBK = methylisobutyl ketone
                                    C-2

-------
    APPENDIX D
ANALYTICAL RESULTS
        D-l

-------
   pcs/lab
   Pollution Control Science, Inc.
   6015 Manning Road, Miamisburg, Ohio 45342
        (513) 866-5908      TLX 288-348
r
~l
           RADIAN CORPORATION
   Report Date:    12-3-80

    Date Rec'd:    10-27-80

  PCS Sampled: .03780 thru 0381

   Client P. OJ:    37064

lnv#/Acct#/PN:    150-1
                                                               ACCREDITED BY THE AMERICAN INDUSTRIAL
                                                                  HYGIENE A SSOCIA TION
                                                               CERTIFIED BY THE OHIO EPA FOR DRINKING
                                                                  WATER ANALYSES
                                                               MEMBER OF THE AMERICAN COUNCIL OF
                                                                  INDEPENDENT LABQRA TORIES
                                         Page  1 of 2
n
i
a
J

[ •
^

1

P-J
J

[]

1
J

:J


a

r]

PLANT M PLATE OVENS 1 AND 2
TOTAL GASEOUS NONMETHANE ORGANICS (TGNMO)
PCS EMISSION
Saitple SOURCE SAMPLE TRAP TANK TOTAL TOTAL
Number
03801
03803 *
03802 *
03798
03799
03804
03805
03800
03796
03797
03813
03808 *
03809 *
03815
03814
03811
03810
03812
03807
03806
* Duplicate
ID
2 -01-1
2 -OI-2A
2 -OI-2B
2 -01-3
2 -01-4
2 -00-1
2 -00-3
2 -00-4
2 --CZ-1
2 -CZ-3
1 -01 -1
1 -OI-2A
1 -OI-2B
1 -01-3
1 -01-4
1 -00-1
1 -00-3
1 -00-4
1 -CZ-1
1 -CZ-3
Samples
VOLUME
2.762
3.148
2.730
3.513
2.503
3.089
3.586
2.184
4.045
3.323
3.431
3.333
3.169
2.829
2.803
3.267
3.778
3.281
4.746
3.568

ppm Cl
2
1
3
1
3
1
1
1


2
2
2
1
3
2

2



J PROCEDURES IN A CCORDANCE WITH 40 CRFPart
~1
October} 6, 1973 138 FR 287581,
and amendments.

739
790
129
856
192
314
345
821
109
95.3
342
754
470
Oil
823
238
86.1
846
69.5
167

736

ppm Cl
10.8
530
17.2
9.5
207
27.3
18.4
27.3
<5
82.3
170
29.1
218
58.7
204
48.2
53.8
250
<5
13.0

^
v. ^x
^
ppm Ci mg/L CL
2
2
3
1
3
1
1
1


2
2
2
1
4
2

3



^
(I

750
320
146
866
399
341
363
848
109
178
512
783
688
070
027
286
139.9
096
69.5
180

• s
4^J j
^ '
1.373
1.158
1.571
0.392
1.697
0.670
0.681
0.923
0.054
0.089
1.254
1.390
1.342
0.534
2.011
1.142
0.070
1.546
0.035
0.090
^^^i "i^
l^H

-------
pcs/Iab
12-3-80
RADIAN CORPORATION
  Page 2 of 2
                                                               PLANT M
                                    01
                 OVEN
               DESIGNATION
SAMPLING
LOCATION
                                                PERIOD
                            SAMPLING LOCATIONS:

                            01 - Oxidizer Inlet

                            00 - Oxidizer Outlet

                            CZ - Cooling Zone
                                     D-3

-------
-I
     pcs/tab
     Pollution Control Science,  Inc.
     6015 Manning Road, Miamisburg, Ohio 45342
          (513) 866-5908      TLX 288-348
  r
  u
                                               Report Date:

                                                Date Rec'd:

                                             PCS Sample*:

                                               Client P.0.#:
         RADIAN CORPORATION
                                    J
                                                   12-3-80

                                                   10-27-80

                                                03780  thru 03

                                                   37064

                                                   150-1
                               ACCREDITED BY THE AMERICAN INDUSTRIAL
                                 HYGIENE ASSOCIA TION
                               CERTIFIED B Y THE OHIO EPA FOR DRINKING
                                 WATER ANALYSES
                               MEMBER OF THE AMERICAN COUNCIL OF
                                 INDEPENDENT LABORA TORIES
                                          Page 1 of 2
t
I
j
•T
,J
1
IM
]
]
1
]

]
PLANT L
PLATE OVEN No. 2
TOTAL GASEOUS NONMETHANE ORGANICS
PCS
Sanple
Number
03794
03791
03790
03780
03781
03795
03787
03786
03785
03792
03788
03789
03793
03782
03783
EMISSION
SOURCE
ID
2-01-1
* 2 -OI-2A
* 2-OI-2B
2 -01-3
2 -01-4
2 -00-1
* 2 -00-3A
* 2 -00-3B
2 -00-4
2 -CZ/A-1
* 2 -CZ/A-3A
* 2 -CZ/A-3B
2 -CZ/B-1
* 2 -CZ/B-3A
* 2 -CZ/B-3B
SAMPLE
VOLUME
3.458
3.016
1.819
2.930
2.219
2.745
3.202
3.607
3.176
2.629
2.524
3.093
3.150
3.064
2.938
TRAP
ppm Cl
1 676
1 627
1 391
554
1 494
208
571
406
127
82.1
155
123
108
435
161
TANK '
ppm Cl
110
69.4
136
101
79
74
42.1
47.9
81.2
139
104
57.7
117
84
102
(TOMO)
TOTAL
ppm Cl
1 786
1 696
1 527
655
1 573
282
613
454
208
221
259
180
225
519
263
TOTAL
mg/L CL
0.892
0.874
0.763
0.327
0.786
0.141
0.306
0.227
0.104
0.110
0.130
0.090
0.112
0.259
0.131
TRAP
No.
7
46
85
35
53
102
101
51
31
24
13
95
57
84
61
TANK
No.
60
53
52
68
69
70
44
71
64
48
49
74
72
56
73
         03784
-FS-1
2.894
210
28.6
239
0.119
89
51
1
            Duplicate Samples
         PROCEDURES IN ACCORDANCE WITH 40 CRF Part 136
         October 16, 1973 138 FR 28758), and amendments.
                                                    D-4
                       aino/i HAS*in is Th« ronfiH^ntial nroo«rtv of Our cHents. Disclosure of such data statement o» conclusion reQuires wnnen approval

-------
pcs/lab
    12-3-80                                                      PLANT L
    RADIAN CORPORATION                 Page 2 of 2
                                    01
                    OVEN         SAMPLING       SAMPLING
                 DESIGNATION     LOCATION        PERIOD
                            SAMPLING LOCATIONS:

                            01 - Oxidizer Inlet

                            00 - Oxidizer Outlet

                            CZ/A - Cooling Zone A

                            CZ/B - Cooling Zone B

                            FS - Floor Sweep
                                         D-5

-------
                                           REPORT
TRUESDAIL LABORATORIES,  INC.
CHEMISTS - MICROBIQLOGISTS  -  ENGINEERS
RESEARCH     -    DEVELOPMENT    -    TESTINS
CLIENT
SAMPLE
Radian Corporation
Post Office  Box 9948
Austin, Texas  78766
ATTN:  Mr. Bill Stadig
20 - hydrocarbon samples collected by USEPA Method
25 and 10  -  paint distillate samples.
P. 0. No.  37066
                                                              4101 N.  FIGUERQA  STF
                                                              LOS  ANGELES  9 O C
                                                              AREA CODE 213  •  225-
                                                              CABLE:   TRUELA
DATE    January  13,  1981

RECEIVED  October 14, 1981

LABORATORY NO.   34533
INVESTIGATION
           Analysis of hydrocarbon samples by USEPA Method 25 and carbon
           determination on  paint distillate samples.
                                            RESULTS

                Evacuated 8 liter tanks and stainless steel freezeout traps were sent
           by air freight to Tom Nelson in La Grange, Illinois.   The returned  samples
           •were analyzed by a Total  Combustion Analysis Method (equivalent to  USEPA
           Method 25 in South Coast  Air Quality Management  District).

                Total of 12 samples  were taken into  8 liter tanks and 8 samples  were
           taken with 6.8 liter  aluminum tanks.
           tent.
                The submitted paint  distillate samples were  also analyzed for  carbon con-
                The results are given in the following tables.
    This report applies only to the sample, or samples, investigated and is not necessarily indicative of the quality or condition of apparently
    identical or similar products. As a mutual protection to clients, the public and these Laboratories, this report is submitted and accepted
    for the exclusive use of the client to whom it is addressed and upon the condition that it is not to be used, in whole or in part, in any
    advertising or publicity matter without prior written authorization from these Laboratories.


                                                      D-6

-------
TRUESDAIL LABORATORIES. INC.
-2-
Laboratory No.   34533
                              Summary of  the Results
                                                     Hydrocarbons
[Plant]
Carbon Dioxide Carbon Monoxide Methane
Sample
R-l
R-2
R-3
R-4
R-5
R-6
R-7
R-10
T-l
T-2
T-3
T-4
T-5
T-6
T-7
T-8
T-9
T-10
T-ll
T-12

Cl
Cl
Cl
C2
Cl
C2
C2
Cl
D
D
D
D
D
D
E
E
E
E
E
E
(CO,,1) , ppm
£.
8900
7000
8800
24,600
7400
23,500
25,000
8100
" 515
416
2483
1006
NR
NR
423
2319
3976
3140
2798
2320
(CO) , ppm

936
981
992
4160
1033
4100
3900
1014
ND
ND
ND
ND
ND
ND
< 15
< 10
< 10
< 10
< 10
< 10
(CHA) , ppm

30
27
33
22
28
21
23
39
ND
ND
ND
ND
ND
ND
< 15
< 10
< 10
< 10
< 10
< 10
Gaseous
(as Ci) , ppm

2600
5400
5600
283
4800
639
105
6300
487
220
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Condensable
(as CO » ppm
11,020
9,031
8,488
314
10,490
512
739
7,595
1531
1809
2203
1214
919
1767
153
120
58
90
42
159
Total
(as Ci) , ppm

13,620
14,431
14,088
597
15,290
1151
844
13,895
2018
2029
2203
1214
919
1767
153
120
58
90
42
159
  Cl = Incinerator Inlet, Plant C
  C2 = Incinerator Outlet, Plant C
  D  = Absorber Outlet, Plant D
  E  = Absorber Outlet, Plant E
  ND = None detected
  NR = Not reported
                                          D-7

-------
TRUESOAIL LABORATORIES. INC.
-3-
Laboratory No.   34533
          Carbon content of solvents (paint distillates)
                 Sample
                  No.

                    Received on 11/18/80

                   33,
                   14
                   18

                    Received on 12/1/80

                    1
                    7
                    9
                   11
                   12
                   15
                   16
        Plant
          D
          D
          E
          M
          L
          L
          D
          D
          D
          D
     % (by weight)
       Carbon (C)
        75.51
        86.85
        74.82
        74.28
        84.83
        85.50
        89.86 (sample appeared
        88.73    cloudy)
        87.03
        88.58 (non-homogeneous
                material  in the
                 sample)
                                         D-8

-------
TRUESDAIL LABORATORIES. INC.
-4-
Laboratory No.   34533
                                     Gas  Tanks
     R-l
     R-2
     R-3
     R-4
     R-5
     R-6
     R-7
     R-10
Volume
Liters
7.75
8.20
7.67
8.23
8.24
8.21
8.083
8.138
8.052
7.905
8.12
8.11
Aluminum
6.80
6.80
6.80
6.80
6.80
6.80
6.80
6.80
"Hg
Res. Vac.
-6.0
-7.7
-20.1
-5.6
-6.4
-14.6
-14.0
-10.0
-11.8
-9.4
-10.4
-12.2
Tanks*
-11.4
-7.6
-7.2
-8.6
-5.8
-8.6
-15.2
-6.6
                                                      80
                                                      72
               72
               74
               74
               72
               72
               74
               74
               74
                           Barometric
                         Pressure "Hg  (corr.)

                               29.53
                               29.40
               29.40
               29.60
               29.60
               29.40
               29.60
               29.60
               29.60
               29.60
     *Two aluminum tanks were standardized for volume.  Both  came out to be 6.80
      "t 0.003 liters.  Therefore 6.80 L.  volume was used  for  the rest of the aluminum tanks.

          The traps were matched with gas tanks as marked.

                                             Respectfully submitted,

                                             TRUESDAIL LABORATORIES, INC.
                                                   Bandziulis
                                             Supervisor, Air Pollution Testing
                                             D-9

-------
               APPENDIX E




MATERIAL BALANCE CALCULATIONS FOR PLANT L
                   E-l

-------
                 MATERIAL BALANCE CALCULATIONS FOR PLANT L


Solvent Use at the Applicator (as CQ


     Ib VOC as Ci  =  105 sheets x  33 125" x 33.369"
          min              min           sheet

   x	5 mg/in2	  x  0.2269 Ib solvent  x 0.8483 Ib Ci	
     1000 mg/g x 454 gm/lb            Ib coating           Ib solvent

                            = 0.246 Ib Ci/min


Afterburner Outlet

                   ppm as Ci                  gas flowrate

                      282
                      613
                      454                        8,120
                      208                        7,830
                  ave 389                        7,975 scfm

           Ib C   =  389  x  12.01 Ib/lb mole  x 7975 scfm        x  0.958
            min      105                          387 scf/lb mole

                            = 0.0922
Afterburner Inlet
                   ppm as GI

                     1786
                     1696
                     1527
                      655
                     1573
                     1447

The incinerator inlet flowrate is assumed equal to outlet  flowrate  (on a
dry basis).  There is no excess air added to the  incinerator.

         Ib d  =  1447  x  12.01_lb  x  7975  scfm  x   0.958  dry
          min       106         Ib mole   387  scf/lb mole     wet

                            = 0.343 Ib Ci/min
                                     E-2

-------
Cooling Zone A
                  ppm as d , dry                flowrate

                       221                       13,340
                       259                       13,310
                       180
                   ave 220                   ave 13,330

     lb Ci  =  220  x!2.011b    x  13,330 scfm     x  0.986 dry scf
     min       10^       lb mole     387 scf/lb mole          wet scf

                             = 0.0897 lb Ci/min


Cooling Zone B

                  ppm as Ci, dry                flowrate

                       225                       10,810
                       519                       10,780
                       263
                   ave 336                   ave 10,800

     lb Ci  =  336  x  12.01 .lb  x  10,800 scfm      x  0.986 dry scf
      min      10s        lb mole   387 scf/lb mole           wet

                             = 0.111 lb Ci/min
Summary of Material Balance

   	lb C /min	
   Applied   Incinerator   Incinerator   Cooling    Cooling    %        %
   Coating	Inlet	Outlet	Zone A	Zone B  Capture  Destrn

    0.246       0.343        0.0922       0.0897     0.111    221      73
                                     E-3

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA 905/2-80-005_
                                                           3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
  Determination  of  Capture and Destruction  Efficiencies
  of Selected Volatile Organic Compound Control Devices
  in the State of  Illinois
             5. REPORT DATE
               December  21,  1980
             6. PERFORMING ORGANIZATION CODE
7. AUTHORfSl
                                                           8. PERFORMING ORGANIZATION REPORT NO.
                                                            DCN 81-240-016-03-09
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  RADIAN CORPORATION
  8501 Mo-Pac  Boulevard
  P. 0. Box 9948
  Austin, Texas   78766      	
                                                            10. PROGRAM ELEMENT NO.
              11. CONTRACT/GRANT NO.
               68-02-3513
               Work Assignment 3
12. SPONSORING AGENCY NAME AND ADDRESS
   U.S.  Environmental Protection Agency,  Region V
   Air Programs Branch
   230 South Dearborn Street
   Chicago,  Illinois  60604
              13. TYPE OF REPORT AND PERIOD COVERED
               Final  -  Aug-Dec 1980	
              14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
  Project  Officer:   Mr. Barry A. Perlmutter
16. ABSTRACT
This report  provides technical support  for the development of  the  Illinois State
Implementation Plan for surface coating industries, and more specifically, paper and
can coating.   A source testing program  was conducted at three  paper coating and two can
coating  facilities in Illinois to  determine the efficiency of  capture and destruction
of volatile  organic compounds (VOC)  using either carbon adsorption or afterburner sys-
tems.  At  the paper coaters, the VOC collection efficiencies were  91-94%, but at the
can coating  plants, collection efficiency was undetermined.  On  the paper coating lines,
two carbon adsorbers showed 79 and 98%  control efficiency and  a  thermal afterburner was
performing at 95% efficiency.  The three afterburners at the can coating plants were con
trolling only 26 to 73% of the VOC's because operating temperatures were relatively low.
EPA Method 25 was used to determine  the VOC concentration in the vapor streams.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS  c.  COS ATI field/Group
  Air Pollution
  Afterburner
  Carbon Adsorption
  Emissions
  Air Pollution Controls
  Volatile Organic Com-
    pounds
  Capture Efficiency
  Destruction Efficiency
       13B
18 DISTRIBUTION STATEMENT

  Unlimited
19 SECURITY CLASS IThisReportl
 None
21. \O. OF PAGES
   114
                                              I 20 SECUR'TY CLASS 'Tins page/
                                              !  None
                           22. PRICE
SPA Fom 2220-1 (Rev. 4-77)   PREVIOUS EDITION 
-------