&EPA
           United States
           Environmental Protection
           Agency
           Office of Air Quality
           Planning and Standards
           Research Triangle Park NC 27711
EMB Report 79-ISC-10
           Air
Industrial Surface
Coating
(Coil)
           Emission Test Report
           Precoat Metals
           Incorporated
           St. Louis, Missouri

-------
EMISSION TEST OF A COIL COATING PLANT IN ST.  LOUIS,  MISSOURI


                             by


                      George W. Scheil
                        October 1980
                        FINAL REPORT
                 EPA Contract No. 68-02-2814
                   Work Assignment No.  28

                  EPA Project No. 79-ISC-10
                  MRI Project No. 4468-L(28)
                             For
                 Emission Measurement Branch
                    Field Testing Section
               Environmental Protection Agency
        Research Triangle Park,  North Carolina  27711

-------
                                    PREFACE
     The work reported herein was conducted by Midwest Research Institute under
Environmental Protection Agency Contract No. 68-02-2814, Work Assignment No. 28.

     The project was under the supervision of Mr. Douglas E. Fiscus, Head,
Field Programs Section and Dr. Ken Wilcox, Program Manager.  Dr. George Scheil
was field team leader and was assisted in the field by Messrs. Jeff Thomas,
Tom Walker, and Charles Brown.  The volatile organic, carbon lab analyses was
performed by Pollution Control Sciences.
Approved for:

MIDWEST RESEARCH INSTITUTE
M. P. Schrag, Director
Environmental Systems Department

October 8, 1980
                                      iii

-------
                                   CONTENTS
Figures	      vi
Tables	     vii

     1.  Introduction	       1
     2.  Summary and Discussion of Results 	       2
     3.  Process Description and Operation 	      15
     4.  Location of Sample Points 	      17
     5.  Sampling and Analytical Procedures	      23

Appendices

     A.  Tentative EPA Method 24 for Volatile Organics Measurements
           of Paint	      27
     B.  Tentative EPA Method 25 for Volatile Organic Carbon 	      33
     C.  VOC Sampling and Analysis Data	      63
     D.  Velocity Traverse Data	     107
     E.  Gas Composition Log Data. .	     117
     F.  Paint Sampling Data Sheets	     137
     G.  Sample Calculations 	     165
                                       v

-------
                                 FIGURES

Number                                                                Page
  1     Time chart for finish coater, afterburner temperature
          1400°F . .	
  2     Time chart for finish coater,  afterburner temperature
          1200°F	       4

  3     Time chart for finish coater,  afterburner temperature
          900°F	       5

  4     Time chart for prime coater,  afterburner temperature
          1400°F 	       6

  5     Time chart for prime coater,  afterburner temperature
          1200°F 	       7

  6     Time chart for prime coater,  afterburner temperature
          1000°F	       8

  7     Process diagram - St. Louis,  Missouri	      18

  8     Afterburner inlet ducts (top  view)  	      19

  9     Boiler/bypass outlet (end view)	      20

 10     Stack (end view)	      21
                                    vi

-------
                                TABLES




Number                                                             Page




  1   VOC/Continuous FID Results	    9




  2   Gas Composition and Flow Rates	   10




  3   Paint Sample Analyses • 	   11




  4   Material Balance Data	   13




  5   Quench Return Duct Data	•	   14
                                  vii

-------
                                   SECTION 1

                                 INTRODUCTION
     This report presents the results of source testing performed during the
period August 28 to September 6, 1979, at the Precoat Metals coil coating plant
in St. Louis, Missouri.  Midwest Research Institute (MRI), U.S. Environmental
Protection Agency (EPA), and Research Triangle Institute (RTI) participated in
the field testing and Pollution Control Sciences (PCS) performed laboratory
analyses for volatile organic carbon (VOC) by tentative EPA Methods 24 and 25
(see Appendices A and B).  Separate test series were run at the prime and finish
coat ovens and their incinerators.  Liquid samples were obtained from the paint
being applied, and gas samples were obtained at the inlet and outlet of the in-
cinerator for each test.

     Sampling included a material balance of volatile organics in the coil coat-
ing process, continuous nitrogen oxides (NOX) and total hydrocarbons (THC) mea-
surements at the process incinerator outlets, volumetric flow measurements at
the inlets and outlets of the incinerators, VOC measurements at the finish oven
quench area, and simultaneous VOC measurements at the inlet and outlet of each
incinerator.

     The results of these tests are to be evaluated by EPA as part of the develop-
ment of emission standards for this industry-

-------
                                    SECTION 2

                        SUMMARY AND  DISCUSSION OF RESULTS
     Figures  1 through 6 show the time sequences for the sampling program.  Refer
to these figures  for correlations between the VOC/continuous monitoring runs and
the material  balance data which follow.

     Table 1  lists  the results of the VOC sampling at the inlet and outlet of
the afterburners  and the continuous flame ionization detector (FID) measurements
of total hydrocarbons at the outlets.  Process upsets occurred during five of the
runs.  Generally, sampling was halted at the upset if at least 20 min of sampling
had occurred.  If less than 20 min had elapsed the run was restarted when the
line stabilized after the end of the upset.  A shortage of sample traps did not
allow the runs to be repeated.  The interruptions seem to have seriously affected
only Runs 2 and 18.  After rejecting these two runs serious inconsistencies re-
main, especially at the lower concentrations, between the THC and VOC results.
The accuracy  of the manually integrated THC traces is limited during Runs 4 and
6 and 16 through  18 due to rapid variations in the THC concentration.  Some heavy
tar buildup was also observed in the THC heated line.  This tar could not have
reached the VOC tanks, which also show a high carbon content.  Data sheets for
the VOC sampling and analysis results are in Appendix C.

     Table 2  contains general information on gas temperatures, composition, and
flow rates.   Data sheets for velocity traverses are in Appendix D.  Log book en-
tries for compositions are in Appendix E.  NOX measurements at the outlet were
by continuous monitor.  The final traverse runs on 9/6 indicate that the flow
and composition at the stack is a reasonable composite of the finish and prime
ducts with little air leakage.

     Moisture for the final traverse at the finish inlet was the average of
Runs 7 through 9.   The prime inlet moisture was the value from Run 18.  Stack
moisture was assumed to be the average of the prime and finish outlet moisture
for Runs 7 and 9 and 16 through 18.

     Table 3 shows the results of analyses of selected paint samples for density
and percent volatiles.  Sampling data sheets and analysis results are given in
Appendix F.

-------
        1000
         I—
TGNMO
Runs
 1100
—I	

 Run #1
                                1200
                               —I—
 1300
—I—
1400
 1500
—I—
1600
1700
1800
                                                               Run#2
                                                                                             Run #3
Process
Down
Times
Material
Balance
Samples
1
Sample *
Top 101*
Back 104
1
Sample *
103
102*
                                                               I
                                                           Sample
                                                              ]Q5~
                                                              106
                              Sample
                                f07~
                                108
                                                                                            1
                                                                                        Sample
                                                                              Sample
                                                                                           109
                                                                                           110
                                                                                 111 *
                                                                                 112 *
Paint
Drum
Changes
                                         Top
                                         Back
                                                             Top
                                                             Back
    *  Sample Analysis  in Table 3.
   _g/Only Process Down Times Which Occurred During  Sampling Runs Are Noted.
                  Figure 1.  Time chart for finish coater, afterburner temperature 1400°F.

-------
      0900
        I
1000
1100
1200
1300
1400
1500
1600
1700
1800
TGNMO
Runs
                  Run#4
                     Run #5
                                           Run#6
Process
Down
Times I
Material
Balance
Samples

Sample "
Top 113
Back 114
1
r I H
Sample v
115
116
\
Sample *
117
118
i
Sample *
119
120
                                       Sample
                                                                   1
                                                                  Sample
                                               Sample
                                                            123
                                                            124
                                                  121 *Tote
                                                  122* Tote
                                                 126
                                                           I
                                                        Sample *
                                                           127
                                                           128
                                                           129 * Top Tote
                                                           130 * Back Tote
                                                           131   Solvent
Paint
Drum
Changes
                                          Top
                                          Back
    *  Sample  Analysis in  Table 3.
    a/ Only Process Down Times Which Occurred During Sampling Runs Are Noted.
                       Figure 2.  Time chart for finish coater, afterburner  temperature 1200°F.

-------
0900
  I
                       1000
           1100
1200
1300
1400
1500
1600
1700
1800
TGNMOI
Runs    I
                                       Run#8
                                                          Run
Process
Down
Times £/
Material
Balance
Samples
1
Sample *
Top 133
Back 134
135 *
Totes 136 #
II
if
Sample '
137
138

' Sample
140
1
# Sample *
141
142
143*
144
1
t #
Sample n
\ 145
146
147
148
1 1
Sample * Sample *
149*
150
151*
152
153
154
155 *
156 *
157 Solvent
Paint
Drum
Changes
Top
                              Top
     *  Sample Analysis  in Table 3.
    _a/Only Process Down Times Which Occurred During Sampling  Runs Are  Noted.
                          Figure 3.   Time chart for finish coater,  afterburner temperature 900°F.

-------
0900
  I
                   1000
1100
1200
 1300
1400
1500
             1600
1700
1800
TGNMO
Runs
                                                Run ^10
                                                     Run * 11
                                                           Run ^12
Process
Down
Times £/
Material
Balance
Samples
               1        1
            Sample ^  Sample
       Top
       Back
  159*
  160
161
162
   I
Sample
   163
   164
            I
         Sample
           165
           166
                                                      Sample *  Sample
                                 167
                                 168
       169
       170
       171 Solvent
Paint
Drum
Changes
              Top
                     Back
                      Top
                      Back
                                   Top
                                                   Back
      *  Sample Analysis in  Table 3.
      a/ Only Process Down Times Which Occurred During Sampling Runs Are Noted.
                        Figure 4.  Time chart for prime coater, afterburner temperature  1400°F.

-------
0900
  I
                    1000
                                1100
                                    1200
                                   1300
                             1400
1500
1600
1700
1800
TGNMO
Runs
                Run*13
Process
Down
Times 2j
Material
Balance
Samples
  Sample *
Top  173 *
Back 174
   I        I
Sample ' Sample
   175      177~
   176      178
Sample
   179~
   180
                                                                i        I
                                                             Sample * Sample ^
                                                                181183
                                                                182       184
                                                                         185   Solvent
Paint
Drum
Changes
                          Top
                          Back
                                             Top
                                             Back
     Sample Analysis in Table 3.
     Only  Process  Down Times Which Occurred During Sampling Runs Are  Noted.
                       Figure 5.  Time chart for prime  coater,  afterburner  temperature 1200°F.

-------
TGNMO
Runs
Process
Down
Times S/
0800 0900 1000 1100 1200 1300
* 1 I I 1 1 1 1 1 1 1 1
Run '16 Run* 17
^^^^^^ ^^^J ^^^^^_
^^1 ^^^^
_ _

                                                                                        1400
                                                                                     1500
                                                                                                               1800
                                                                                                                    Run* 18
00
Material
Balance
Samples
                            I     I
                        Sample * Sample '
Top  187 *',
Back 188  |
189
190
   I
Sample
  J91~
  192
Sample
  193~
  194
Sample
  195~
  196
   I
Sample *
  197
  198
  199 'Solvent
              Paint
              Drum
              Changes
                                                                         Top
                                                                         Back
                  * Sample Analysis in Table 3.
                 _g/Only Process Down Times Which Occurred During Sampling Runs Are  Noted.
                            Figure 6.   Time chart  for  prime coater, afterburner temperature 1000°F.

-------
                                               TABLE 1.   VOC/CONTINUOUS  FID RESULTS
Run
No.
I
2—'
3
4
5
6
7
8
9
10
ni'
12
13
14
15b/
16
11&
18-

PPH C (trap)
16,006
631
6,035
6,711
8,066
8,941
5,734
7,058
9,389
5,466
6,290
6,570
-£'
2,280
1,202
926
634
2,593
Inlet
PPH C (tank)
582
561
3,600
2,121
573
2,846
3,732
3,425
3,242
293
567
405
_»l
534
535
511
459
615

PPH C (total)
16,588
1,192
9,635
8,832
8,639
11,787
9,466
10,483
12,631
5,759
6,857
6,975
—3'
2,814
1,737
1,437
1,093
3,208

PPH C (trap)
841
304
-a/
84
827
295
4,557
3,084
4,571
148
100
153
184
276
76
873
533
347
Outlet
PPH C (tank)
387
174
-a/
183
239
312
678
454
409
123
170
145
384
220
195
591
528
458
TllC-contlnuou
PPH C (total)
1,228
478
-a/
267
1,066
607
5,235
3,538
4,980
271
270
298
568
496
271
1,464
1,061
805
PPH propane

0.4
0.75
11
10
10
2,125
1,850
1,800
-
8.0
9.0
12.5
11.0
10.5
315
350
275
PPM Ci

1.2
2.2
33
30
30
6,375
5,550
5,400
-
24
27
37.5
33.0
31.5
945
1,050
825
s FID-outlet
Remarks

Stable trace
Stable trace
Erratic trace
Erratic trace
Erratic trace
Stable trace
Stable trace
Stable trace

Stable trace
Stable trace
Stable trace
Stable trace
Stable trace
Medium stability trace
Medium stability trace
Medium stability trace
jj/  Trap damaged in shipraent--not analyzed.




J>/  Process upset during run, TGNMO results  are of uncertain validity.

-------
                                                   TABLE  2.   GAS  COMPOSITION AND  FLOW  RATES
Run No
1
2
3
4
5
6
7
8
9
10
11
12

13
14
15
16
17
IS


Incinerator

temperature
. Date/time (°F) Coating X CO,
8/28/1042 - 1132
8/28/1432 - 1502
8/28/1654 - 1733
8/29/0940 - 1025
8/29/1129 - 1215
8/29/1414 - 1500
8/30/0910 - 0953
8/30/1106 - 1152
8/30/1438 - 1522
9/4/1206 - 1251
9/4/1506 - 1525/1539 - 1552
9/4/1636 - 1728

9/5/0922 - 1012
9/5/1053 - 1145
9/5/1334 - 1418
9/6/0830 - 0903
9/6/0954 - 1013/1039 - 1107
9/6/1540 - 1544/1701 - 1729
9/6
9/6
1400 Finish 0.0,
1400
1400
1200
1200
1200
900
900
900
Jtf
2.0
3.0
2.8
3.1
1.6
1.2
1.1
1400 Primer 3.2
1400
1400

1200
1200
1200
1000
1000
1000
3.0
3.1

2.6
2.0
1.9
2.6
1.8
2.2
1000 Finish 3.3
1000 Stack 3.0

7. 02
19. Of*.
. -'
16.2
17.0
18.5
19.5
19.1
18.8
19.0
18.0
16.9
17.4

18.0
17.6
17.0
18.4
18.0
18.7
19.4
16.3

7. II20
7.8
10.1
8.7
6.8
6.9
7.1
7.4
7.9
7.0
10.0
8.3
10.4

9-7
9.9
9.3
9.7
9.9
9.1
7.4
8.7
Inlet
Temperature
702
(372)

700
(371)

532
(278)

715
(379)


656
(347)

716
(380)

621 (327)
935 (502)
Outlet
Flow rate
dsft'/mln (dsnrVrain)

7,870
(223)

8,610
(244)

10,070
(285)

6,540
(184)


6,970
(195)

7,470 (209)

9,260 (260)
20.910 (534)
% C02
5.5
4.6
5.2
4.2
3.8
4.3
3.6
2.2
2.5
5.6
5.5
5.4

5.1
5.2
5.2
3.3
3.9
4.1


7. 02
13.5
13.3
11.1
15.2
14.7
14.8
17.6
17.0
17.8
14.2
13.8
13.4

15.6
15.2
15.3
16.7
17.4
17.4


% II20
11.4
13.2
11.7
9.4
8.6
8.7
7.4
3.8
7.8
11.3
8.0
12.3

10.5
11.5
11.9
10.1
10.4
8.9


Temperature
°F, (°C)
310 (155)
300 (150)
290 (145)
280 (135)
285 (140)
290 (145)
290 (145)
290 (145)
290 (145)
280 (135)
_£/
jJ
c/

--.
JJ
_c/
~cV



HOX
(ppm N02)
34
34
36
17
15
16
7.9
6.7
7.5
29
26
29

37
30
31
27
30
32


aj  Bag leaked, value suspect,  remaining  samples at inlet measuring direct from duct.




b/  Line upset occurred before  sample obtained.




cj  Potentiometer inoperative - temperature not obtainable.

-------
TABLE  3.   PAINT  SAMPLE ANALYSES

Sample
No.
101
102
111
112
121
122
129
130
135
136
143
149
151
155
156
159
173
187

Top coat Run 1 before
Back coat Run 1 after
Top coat Run 3 after
Back coat Run 3 after
Top tote Run 6
Back tote Run 6
Top tote Run 6 - end
Back tote Run 6 - end
Tope tote Run 7 before
Back tote Run 7 before
Top tote Run 8 before
Top coat Run 9 before
Top tote Run 9 before
Top tote Run 9 after
Back tote Run 9 after
Top coat Run 10 before
Top coat Run 13 before
Top coat Run 16 before
% Volatile
("/«)
27.23
31.99
25.75
31.61
26.47
32.86
28.48
34.17
25.24
30.69
25.74
24.81
28.28
27.35
33.48
46.98
46.47
46.16
Density
gin/cm
1.3185
I . 2506
1.3250
1.2519
1.3073
1.2137
1.3098
1.2496
1.3087
1.2247
1.3073
1.2738
1 . 3050
1.3022
1.2179
1.0928
1.1.123
1.1220
% Volatlles
(v/v)
39.94
44.97
37.84
44.29
39.04
45.36
42.19
48.42
37.29
42.69
37.49
35.67
41.69
40.27
46.35
57.87
60.36
55.98
Kg C/f solids
.368
.554
.372
.528
.405
.540
.445
.663
.383
.493
.404
.346
.434
.439
.583
.905
.804
.635
Ib C/gal. paint
1.84
2.54
1.93
2.43
2.06
2.46
1.78
2.85
2.00
2.36
2.11
1.86
2.11
2.19
2.61
3.18
2.66
2.33
                                                                             .2210
                                                                             .3049
                                                                             .2312
                                                                             .294]

                                                                             .2469
                                                                             .2951
                                                                             .2128
                                                                             .3420
                                                                             .2402
                                                                             .2825
                                                                             .2525
                                                                             .2226
                                                                             .2531
                                                                             .2622
                                                                             .3128

                                                                             .3813

                                                                             .3187

                                                                             .2795

-------
     Table 4 summarizes the material balance data.

     Table 5 presents the results of sampling at the finish line quench zone,
The flow rate and low hydrocarbons content indicate that this vent is not a
significant emission source.

     Sample calculations are included in Appendix G.
                                      12

-------
                                      TABLE  4.   MATERIAL BALANCE  DATA
Run Uur (.peed f*-tal wlJtlt H_>i .il ((JUKI-
Nn. ll/n-ln (In.) (in.) H-/«In IB
',/
3
'•
7
a
*
2
.j
4

6
74-'

5" 41-1/16 O.U1
.11 41-9/16 0.02
Ml 41-3/8 0.01
5O 41 • 3/8 O.O1
10 35- .1/8 0.01
5H 35-3/8 0.01
5<) 15-1/8 0.01
50 /.I-0/I6 0.01
50 41-9/16 0.01
50 • 40 0.01
50 40 0.0|
50 40 0.0|
50 36 0.01
50 16 0.01
SO 36 0.01



5
5
5




5
5
5

4111

,03O
,O10
,UJ"
,210
,210
.170
,170
,170
C.iiiit use.t OrrJiiiC lo-idinR-' Inclncrjlor iul.-t Inc lu-r*l 
H
4
I,
1
I
3
I
1
- 1'.' '• .0*. 1 . H4 . ,,«R. 0.11 II . H •*-'.• " I • "
.'•4 7.1h .07 2%. B
.">2 2. It, .07 W u 0.071 ) 2 « lir:> 97.il '"».4

.24 2.45 . 3i nog. I .«, J o. I'. i4 . 7 '('. >
.11 3.1'fl 1 . '.4 nt-K. 1 , IO O.Kl (.'..S
.5'- I.V5 | .71 nep.. I .*i7 O. 71 MI. I
,4'i 1 • (9 O.6 ) 5: 0~- "I. I
.11 0.17 o.l7 8'1-1 0.060 2.7 x 0~y 81-8
.IX. 0. 11 o. 15 8^'H U.1J 0.15 II
.21 0.26 0 t2 90.4 0.24 0.11 7.7
.32 0. 7S „. j4 M . ? 0. 185 O.O84 7S. 3
. 1
. H
. 1
.1
_•}

.R
.1
.*.
iiiul l?2. 110. I3h, 156. Runs 10 through
                                        . 121. 121, Ml. 143. 149. I'M.

-------
 TABLE 5.  QUENCH RETURN DUCT DATA
THC < 15 ppm Cl 21% oxygen
Flow rate - 200-250 ft/min
Flow - 45 ft3/min (1.25
VOC results - 129, 122, 164 ppm Cl
                14

-------
                                   SECTION 3

                       PROCESS DESCRIPTION AND OPERATION
PROCESS DESCRIPTION

     Precoat Metals is a toll coater of metal coil and is owned by Chromalloy.
They operate two coil coating lines in their plant at St. Louis.  Both lines
can process metal in widths from 18 to 44 in. and thicknesses of .008 to .050 in.
In 1978 a B & K Machinery Company system of emission control was installed on the
No. 2 line.  This system uses zone incinerators inside the ovens to burn the sol-
vent emitted from the metal coatings.  A waste gas incinerator is also used with
this system in conjunction with a waste heat boiler that supplies process steam
for the wet section of the coil coating line and supplies steam for building
heat in the winter.  The burning of the solvent by the zone incinerators helps
supply heat for curing the coatings, and by recirculating the air within the ovens,
greatly reduces the amount of dilution air that must be brought into the oven.

     The No. 2 line at this plant is fairly typical of coil coating lines in gen-
eral.  It consists of a decoiling or entry station, a joiner station, an inlet
accumulator, a wet section for metal cleaning and pretreatment, a prime coat sec-
tion with oven and quencher, a finish coat section with oven and quencher,  an
exit accumulator, an inspection station, a metal shear, and a recoil station.
Each of the coating stations is enclosed in a coater room.  A long metal hood
extends from the point on the oven where the metal strip enters to a point very
close to the coating applicator rolls.  Air, supplied into the coating room from
outside, is drawn into the oven through this hood.  The hood extends into the
coating room approximately 14 ft and completely encloses the metal strip over
most of this length.  The hood is for the purpose of capturing any solvent that
flashes off the strip before entering the oven.  Maximum line speed of the No. 2
line is 350 fpm.  The ovens are normally operated at temperatures of 800 to 900°F
to achieve the curing temperature in the metal strip of 420 to 435°F for the
prime coat and 400 to 420°F for the finish coat.

     Most of the metal coated by Precoat is used in the building industry for
siding, roofing, gutters, awnings, etc.  This accounts for about 95 percent of
their total production.  The remaining 5 percent is used for ducting, signs, and
other miscellaneous items.  Practically all of the metal is hot dipped galvanized
steel.  In most instances, the coatings and suppliers are dictated by the custo-
mer.  Types of coatings used are epoxy resins and primers, polyesters, acrylics,

                                      15

-------
fluorocarbons, siliconized acrylics and polyesters, and a small amount .of poly-
vinyl chlorides.  Coatings suppliers include Conchemco, Midland Dexter, PPG,
Mobil, Enterprise, Lily, Whitaker, and others.  Prime coats are generally ap-
plied with a thickness of 0.2 mil on both sides of the metal; the top coat is
usually applied with a thickness of 0.5 mil on the back side and 0.8 mil on the
front side.  Prime coatings are 30 to 32 percent solids by volume, and top coat-
ings are approximately 50 percent solids by volume.

     Some major changes in the No. 2 line have been made since its installation.
These changes involved modifications to the tracking system and the water quench
system.  The line speed has been increased from a maximum of 300 fpm to 350 fpm.

     Solvent input on the No. 2 line amounts to approximately 1,300 gal. every
24 hrs.  Types of solvents used are mostly aromatic hydrocarbons and glycol
ethers such as Solveso 150 and butyl cellosolve.  Reclaimed solvent (mostly ke-
tones) is used for cleanup.

OPERATION DURING TEST

     The x^eather on August 27 to 31 was clear and sunny with temperatures in the
low 90's.  The weather during the period September 3 to 7 was clear and sunny
with temperatures in the high 80Ts, except for September 5, which was cloudy
with a light rain in the mid afternoon.  No effect on the test results is ex-
pected from these weather conditions.

     There were three instances when the tests were interrupted because of ad-
justments that had to be made to the system.  During Run 18, testing was inter-
rupted for 75 min.; a 27 min. interruption took place during Run 17; and a
15 min. interruption took place during Run 11.  In each of these instances sam-
pling was shutdown not more than 2 min. after the interruptive period began and
was reinitiated only after the resumption of coating, followed by a short time
period to ensure the stabilization of the zone temperatures and the afterburner
temperatures.

     The preheat burner was shutdown for the second and third tests of the finish
afterburner exhaust at 900°F.  The preheat burners were shutdown for the three
tests of the prime afterburner exhaust at 1,000°F.

     With the exception of the instances mentioned above there were no major
process or control equipment upsets during the test periods or between the test
periods.  The process was judged to be running at the expected capacity during
the testing.
                                       16

-------
                                   SECTION 4

                           LOCATION OF SAMPLE POINTS
     The process diagram is shown in Figure 7.  Primary sampling occurred at
Points 1, 5, 9, 10, 11, 12, and 13.  Points 9 and 10 (afterburner inlets) are
shown in Figure 8.  Points 11 and 12 (afterburner outlets) are shown in Figure 9,
and Point 13 (stack) is shown in Figure 10.  Sampling was also done at Point 6.
Points 3, 4, 7, and 8 had poor accessibility and were not sampled.

     The primary and finish coat sections were tested separately.  Thus, Runs 10
through 18 were at Points 1, 9, and 11.  Runs 1 through 9 were at Points 2, 10,
and 12."

     One test included:

     Points 1 or 2

     1.  Grab sample(s) of paint (2), each side coated independently.
     2.  Paint usage by weight difference.
     3.  Coil area processed during time limits of step (2).
     4.  Process data information and test start scheduling.
     5.  Pain coating characteristics (thickness, adhesion, solids content, etc.).

     Points 9 or 10

     1.  Velocity - single point.
     2.  H20, C02, 02, temperature.
     3.  TGNMO
     4.  Volumetric flow rate.

     Points 11 or 12

     1.  C02, 02, temperature.
     2.  TGNMO
     3.  THC
     4.  Continuous NOX-
                                       17

-------
00
            Prime
            Coater
           Finish
           Coater
Zone
Incinerator
Prime Oven

Zone
Incinerator
                                      Finish Oven
                                                                            Quench
                                                                            Area
                                                                            Quench
                                                                            Area
Afterburners and
Waste Heat Boilers
                                                                                                                      Stack
                                     Figure 7»   Process diagram -  St. Louis,  Missouri.

-------
                     X
                     I
Duct  I.D.'s are
30" High,  36" Wide
        4 Ports
        on Vertical
        1" Minimum
        Clear  Bore  ( 9
        Scalable
       Blowers,
       Afterburners
       and Boiler
            About
            7' High
                                 4 Ports on  (10
                                 Vertical
                                                              Finish
4'x4'
                                                              Prime
•4'x4'
                                    •Expansion
                                     Joints
                                     7'
                                  Minimum
                  Figure  8.   Afterburner  inlet ducts  (top  view)
                                        19

-------
to
o
                                                                                           Outlet

                                                                                           Sampling
                                                                                       Scaffold at  End
                                    Figure 9.  Boiler/bypass outlet  (end view).

-------
3" Pipe Ports
                      \
Boiler
Primer
Afterburner
Finish
                                                          /
/
                                                                              2 Scaffold
5
                                                                                   Parking
                                                                                   Lot
                                                                                 Alleyway
                              Figure 10.  Stack  (end view).
                                             21

-------
     There was also sampling at the finish quench area—continous FID measure-
ments—flow rate, and VOC—for the final test day.

     One set of velocity traverses was completed at Points 9, 10, and 13, in-
cluding 02, C02, and temperature to confirm that the flow and composition out
the stack was a reasonable composite of the flows going into the incinerator.
                                      22

-------
                                   SECTION 5

                      SAMPLING AND ANALYTICAL PROCEDURES
     The sampling equipment for VOC was prepared by PCS.  MRI personnel con-
ducted the sampling and the sample trains were returned to PCS for analysis
according to Method 25 (see Appendix B) .

     The cold trap inlet lines were too short to reach into the ductwork.  MRI
had to attach 2 ft extensions to obtain enough length for sampling.  The exten-
sions had to be reused, but inlet and outlet extensions were not exchanged, nor
do the trap portions of the samples show any pattern of buildup.  The high levels
reported also appear in the tanks, which could not be affected by any compounds
remaining in the extensions.  The extensions were attached during leak checks so
that any volatiles remaining would have gone to the vacuum pump.

     The line upsets may have affected some samples since it usually took 1 to 3
min. for the sample trains to be shut off after an interruption.  The control
valve was closed but the probe not capped during the interruption which isolated
the tank but may have allowed some transfer to or from the trap.

     The VOC analytical procedure used by PCS followed the EPA proposed Method 25
except in the calibration procedure and catalyst checks:

     1.  Calibration of the analyzer for the analysis of the combusted trap con-
tents is performed at the following conditions:

     a.  Oxidation catalyst - on-line
     b.  Reduction catalyst - on-line
     c.  Column - 100°C

An attenuation is chosen based on estimated concentrations from the trap burnout
traces (NDIR output) and triplicate injections of two or three standards (C02 in
air) are made.  Triplicate injections of the intermediate collection tanks are
then made and concentrations calculated by comparing peak areas to the best fit
straight line of the standard data.
                                       23

-------
      2.  Calibration  for  the analysis of the  tank portion of  the sample is done
again using standards chosen to bracket the expected range of the samples being
analyzed.  An attenuation is chosen on the FID to provide adequate sensitivity
and two or three calibration standards are injected in triplicate.  Peak areas
are measured by an electronic integrator and  the best fit straight line is cal-
culated for the resulting area versus concentration data.  From this,  the sample
concentrations are calculated for the nonmethane organics backflush peak.  This
calibration procedure is  done at a minimum before and after analysis of a set of
samples.  Recalibration is, of course, done should any of the samples  require a
sensitivity change.

      3.  The oxidation catalyst efficiency check is made at the following condi-
tions:

      a.  Reduction catalyst - bypassed
      b.  Oxidation catalyst - on-line at 860  ± 20°C
      c.  Column - either  at 0°C or 100°C

Injections of a standard  mixture of CIfy are made at maximum sensitivity and any
response noted.  If oxidation is 100 percent  no response will show up.  If a re-
sponse is noted, the concentration is measured and an efficiency of oxidation
calculated.  An average efficiency of 99.5 percent or greater for triplicate in-
jections is judged acceptable.

      4.  The reduction catalyst check is performed as follows:

      a.  Reduction catalyst - on-line at approximately 400°C.
      b.  Oxidation catalyst - bypassed.
      c.  Column - 0°C to  permit separation of CO  and CH^.

Injections of a mixture of equal concentrations C02 and CH4 are made and the
resulting peak areas compared.  Efficiencies  typically are 99  to 100 percent,
which is considered adequate since the manufacturers analysis  of the standard
mixture is accurate to only + 2 percent.

     A heated sample line was used for the continuous NOx and  THC analyzers.
A Tee fitting placed at the inlet to the sample conditioning manifold split off
the flow to the Beckman model 402 THC analyzer, which has its  own heated line,
oven, stainless steel pump and flame ionization detector.  The remaining sample
passes through a particulate filter, a Permapure extractive dryer,  Teflon-coated
diaphragm pump and finally one of a pair of calibrated rotameters for the neces-
sary analyzer dilution within the continous monitor manifold  system.   NOx mea-
surements at the outlet were made using a Bendix 8101B chemiluminescent analyzer
set to the total (NO + N02)  mode.  Since this instrument has a highest range of
5 ppm the sample gas was diluted with prepurified nitrogen for analysis.  Cali-
bration was with a 150 ppm NO in nitrogen gas for NOX and with a 99 ppm propane
in nitrogen gas mixture for THC.  Both continuous analyzers have linear re-
sponses (within 10 percent)  and are normally stable over several hours within
the same limits.
                                      24

-------
     Velocity was determined using EPA Method 2 once each day at the inlet and
on the final day at both inlets and at the common stack.  Stack temperature at
other times was too high to complete traverses safely'due to radiation from the
unlined stack.

     C02 and 02 at the outlet were measured by Fyrite from an integrated gas bag.
The ambient temperature at the inlet caused failure of the Tedlar bags so that
sampling was conducted directly from a line connected with the duct port.  After
analyzing the sample from Run 16-Inlet, the oxygen measuring Fyrite was recharged
and a gasket was found to have been installed backwards, which caused all previous
02 readings to  indicate 1 percent high.

     Moisture was measured using EPA Method 4 (midget impinger train).  The mois-
ture readings during the last nine runs were low due to the theft of the balance
used to weigh the silica gel impinger.  The moisture content for Runs 10 through
18 was corrected by adding 0.6 percent H20 to the measured values (saturated H20
vapor at 0°C).

     Paint samples for material balance were obtained in one pint paint cans
direct from the paint drums.  Selected samples were analyzed by PCS according
to tentative EPA Method 24 (see Appendix A).  Paint consumption was determined
with a measuring stick inserted into each drum of paint during the run.  Varia-
tions in level of the paint trays and frequent use of return drums made the paint
consumption difficult to determine during the tests.  As a result, some of the
run to run data scatter shown for the results are expected to be the result of
paint volume measurement errors and not true variations in paint consumption.
                                       25

-------