&EPA
          United States
          Environmental Protection
          Agency
         Office of Air Quality
         Planning and Standards
         Research Triangle Park NC 27711
EMB Report 79-RBM-3
          Air
Rubber Products
Styrene-Butadiene
Rubber Manufacture

Emission Test Report
Phillips Chemical
Company
Rubber Chemicals-
Copolymer Plant
Borger, Texas

-------
          SOURCE TEST AT PHILLIPS
    PETROLEUM'S STYRENE BUTADIENE RUBBER
                  PLANT

               Borger, Texas

          Contract No. 68-02-2812
          Work Assignment No. 65
     Technical,Manager:  Terry Harrison
               Prepared for:

United States Environmental  Protection Agency
 Emission Standards and Engineering Division
        Emission Measurement  Branch
Research Triangle Park, North  Carolina  27711
                   By:


                  TRW
  ENVIRONMENTAL ENGINEERING DIVISION
              P. O. Box 13000
   RESEARCH TRIANGLE PARK, N. C. 27709

-------
                            TABLE OF CONTENTS



Section                                                          Page



   1      INTRODUCTION	       1-1



   2      SUMMARY OF RESULTS	       2-1



   3      DISCUSSION OF RESULTS  	       3-1



   4      PROCESS DESCRIPTION	       4-1



   5      SAMPLE LOCATIONS 	  	       5-1



   6      SAMPLING AND ANALYTICAL PROCEDURES ...  	       6-1



   7      METHOD DEVELOPMENT 	       7-1



   8      QUALITY ASSURANCE AND QUALITY CONTROL	       8-1





          APPENDIX A  SAMPLE CALCULATIONS	       A-l



          APPENDIX B  ANALYTICAL RESULTS 	       B-l



          APPENDIX C  FIELD LOG	       C-l



          APPENDIX D  PROCESS DATA	       D-l



          APPENDIX E  FIELD SAMPLING DATA SHEETS  	       E-l



          APPENDIX F  ANALYTICAL PROCEDURES	       F-l



          APPENDIX G  SAMPLING PROCEDURES	       G-l



          APPENDIX H  PROJECT PARTICIPANTS 	       H-l

-------
                             LIST OF TABLES
Number                                                           Page
 2-1      General Summary of Average Emission Data Collected
            at Phillips Petroleum, Borger,  Texas  Facility  . .  .    2-2
 2-2      Test Result Summary of the Solution Crumb Process .  .    2-5
 2-3      Solution Process Steam Stripper Results 	    2-7
 2-4      Test Result Summary of the Emulsion Crumb Process .  .    2-9
 2-5      Styrene Removal  Efficiency of "B"  and "E" Stripper
            Columns in Emulsion Crumb Process . .  .	    2-11
 3-1      Sub-Summary of Results at the Solution  Crumb  Process -
            Flash Tank Inlet	    3-2
 3-2      Sub-Summary of Results at the Solution  Crumb  Process -
            Flash Tank Outlet	    3-3
 3-3      Sub-Summary of Results at the Solution  Crumb  Process -
            Crumb Tank Outlet	    3-6
 3-4      Sub-Summary of Results at the Solution  Crumb  Process -
            Dewatering Screen Hoods 	    3-7
 3-5      Sub-Summary of Results at the Solution  Crumb  Process -
            Cyclone Vent	    3-8
 3-6      Sub-Summary of Results at the Solution  Crumb  Process -
            Process - Dryer Vent A  	    3-10
 3-7      Sub-Summary of Results at the Solution  Crumb  Process -
            Dryer Vent B	    3-11
 3-8      Sub-Summary of Results at the Solution  Crumb  Process -
            Dryer Vent C	    3-12
 3-9      Sub-Summary of Results at the Solution  Crumb  Process -
            Dryer Vent D	    3-13

-------
LIST OF TABLES (continued)
Number
3-10

3-11

3-12

3-13

3-14

3-15

3-16

3-17

3-18

3-19

3-20

3-21

3-22


Sub-Summary of Results at the Solution Crumb Process -
Dryer Vent E 	 	
Sub-Summary of Results at the Emulsion Crumb Process -
Kerosene Knockout on 6/16/80 	
Sub-Summary of Results at the Emulsion Crumb Process -
White Line Roof Vent A 	
Sub-Summary of Results at the Emulsion Crumb Process -
White Line Roof Vent B 	
Sub-Summary of Results at the Emulsion Crumb Process -
White Line Roof Vent C 	
Sub-Summary of Results at the Emulsion Crumb Process -
White Line Roof Vent D 	
Sub-Summary of Results at the Emulsion Crumb Process -
White Line Drier Vents 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Roof Vent B 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Roof Vent C 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Roof Vent D 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Dryer Vent A 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Dryer Vent B 	
Sub-Summary of Results at the Emulsion Crumb Process -
Black Master Line Dryer Vent C 	
Page

3-14

3-16

3-18

3-20

3-21

3-22

3-23

3-25

3-26

3-28

3-29

3-30

3-31

-------
                       LIST OF TABLES (concluded)
Number                                                           Page
 3-23     Sub-Summary of Results at the Emulsion Crumb Process  -
            Black Master Line Dryer Vent D	    3-32
 4-1      Emission/Production Rate	    4-4
 8-1      Degradation Study SDO-A-3 	    8-2
 8-2      Results of Audit Samples From Borger,  Texas  	    8-10

-------
                             LIST OF FIGURES
Figure                                                           Page
 2-1      Emulsion Crumb Process	     2-10
 5-1      Flash Tank Accumulator with Ammonia Condenser:
            Sample Points SFI and SFO	     5-2
 5-2      Sampling Apparatus at SFO	     5-4
 5-3      Blend Tank Ammonia Condenser:   Sample Points  SBI
            and SSI	     5-6
 5-4      Crumb Slurry Tank Outlet:   Sample Point SCO	     5-8
 5-5      Solution Process - SHO; SYO, SDO's	     5-10
 5-6      Dewatering Screen Hood Vent:  Sample Point SHO  ...     5-11
 5-7      Cyclone Vent:   Sample Point SYO	     5-13
 5-8      Solution Crumb Dryer Vents:   Sample Points SDO-A,
            SDO-B, SDO-C, SDO-D, and SDO-E 	     5-14
 5-9      Emulsion Process Kerosene Tank:   Sample Point-EKO.  .     5-15
 5-10     White Line Roof:   Sample Locations ERO's and  EDO  .  .     5-18
 5-11     White Line Processing Building 	     5-19
 5-12     Black Master Roof:  Sample Locations ERO-BM's,
            EDO-BM's	     5-20
 5-13     Black Master Processing Buidling 	     5-21
 5-14     Emulsion Crumb Process Roof Vent:   Sample Points
            ERO-A, ERO-B, ERO-C, ERO-D,  ERO-BMB,  ERO-BMC,
            ERO-BMD	     5-23
 5-15     Emulsion Crumb Process, White Line Dryer Vent:
            Sample Point EDO	     5-24

-------
                       LIST OF FIGURES (continued)
Figure                                                           Page
 5-16     Emulsion Crumb Process Black Line Dryer Vents:
            Sample Points EDO-BMA, EDO-BMB, EDO-BMC,
8-1
8-2
8-3
8-4
8-5
8-6
Benzene Calibration Curve ^nd Audit Results. . . .
Styrene Calibration Curve and Audit Results. . . .
Cyclohexane Calibration Curve and Audit 	
Hydrocarbon Degradation Study 	
Hydrocarbon Degradation Study 	
*
Hvdrocarbon Dearadation Studv 	
, . 8-3
, . 8-4
, . 8-5
. . 8-6
, . 8-7
, . 8-8

-------
                        GLOSSARY OF TERMS

EMB - Emission Measurement Branch
EPA - Environmental Protection Agency
FID - Flame lonization Detector
GC  - Gas Chromatograph
GPM - Gallons per Minute
HC - Hydrocarbons
NESHAP - National Emission Standards for Hazardous Air Pollutants
NSPS - New Source Performance Standards
SBR - Styrene Butadiene Rubber
TBC - Tertiary Butyl Catechol
TC  - Thermal Conductivity
TGNMO - Total Gaseous Non Methane Organics
VOC - Volatile Organic Compounds

                     SAMPLE POINT DESIGNATIONS
                                                                 /
1st letter
E - Emulsion Crumb Process
S - Solution Crumb Process

2nd letter
K - Kerosene Knockout
S - Steam Stripper
D - Drier
R - Roof Vent
B - Blend Tank Ammonia Condenser
F - Flash Tank Ammonia Condenser
C - Crumb Tank Vent
H - Hood Vent
Y - Cyclone

3rd letter
I - Inlet
0 - Outlet

Example:  SDO = Solution Crumb Process Drier Outlet

-------
                          1.0  INTRODUCTION

     Under the supervision of the Emission Measurement Branch (EMB) of
the Environmental Protection Agency (EPA), TRW Environmental Engineering
Division personnel conducted a study of the volatile organic compound (VOC)
emissions produced by the styrene-butadiene rubber (SBR) industry at
Phillips Petroleum's facility in Borger, Texas.
     Information obtained from the April 24, 1980 presurvey of the
Borger plant was used to develop a test plan.  This plan was implemented
through the field sampling and analysis which was performed from May 27
through June 22, 1980.
     The purpose of this test was to obtain information necessary for
the support of possible new source performance standards (NSPS) and the
national emission standards for hazardous air pollutants (NESHAP).   The
main objectives of the testing were:  1) to determine the total volatile
organic compound emissions from the emulsion crumb and solution crumb
processes; 2) to determine efficiencies of the control devices of each
process; and 3) to determine the degradation rates of organic compounds
in the volatile organic samples.  The specific compounds under analysis
include:  butadiene, cyclohexane, C,-Cg alkanes, benzene, toluene,
ethylbenzene, xylenes, styrene, carbon dioxide,  oxygen, and nitrogen.

-------
                         2.0  SUMMARY OF RESULTS

     This section presents summary tables of results and narrative on
the emissions tested during the period of May 27 through June 22, 1980
at Phillips Petroleum's Styrene Butadiene Rubber Plant in Borger, Texas.
Testing was performed on various gas and liquid streams along two dif-
ferent process lines at the Borger facility.   The testing of the processes
was coordinated with Phillips Petroleum personnel.   A TRW mobile laboratory
containing sampling and analytical equipment was transported to Borger
for on-site analysis of all gaseous samples.   Liquid samples were prepared
on-site and shipped to TRW's Research Triangle Park facility for analysis.
The testing was performed utilizing standard and proposed EPA reference
methods (see Section 6).
     Hydrocarbon concentrations were calculated from peak area summations
from the gas chromatography/flame ionization detection (GC/FID) analysis
method.  The liquid analysis samples were analyzed by the same GC/FID
analysis except that the liquid samples required further preparation as
described in the EPA draft methods presented in Appendix F.   EPA Method
25 was employed as a comparative method for hydrocarbon determination at
the kerosene knockout sample location (EKO).   Table 3*4 presents the
results of the Method 110 and Method 25 test results at the EKO position.
The inert gas concentrations of the gas samples were analyzed by gas
chromatography/thermal conductivity detectors (GC/TCD).  The three
analytical methods are presented in Appendix F and discussed for adap-
tions at each sample location in Section 6.  The sampling location
positions are discussed in Section 5 and the adaption from standard EPA
procedures are presented in Section 6.
2.1  SUMMARY OF EMISSION RATES DURING TESTING
     Table 2-1 presents an overall emission rate of plant operation
during the test period with a comparison of the emission rate to the

-------
                Table 2-1.    GENERAL SUMMARY  OF AVERAGE  EMISSION  DATA-
                               COLLECTED AT  PHILLIPS  PETROLEUM
                                        BORGER, TEXAS  FACILITY
Sample Point Description
lab No. Hydrocarbon
Emission
Concentration
(PPM at Benzene)
Hydrocarbon
Emission
Rate
(Lbs/Hr)
Hydrocarbon
Emission
Production
Rate
Ubs/Lbt")
SOLUTION CRUMB PROCESS

Crumb Tank Vent Outlet
Hood Vent Outlet0
Cyclone Outlet


Dryer Outlet A
Dryer Outlet B
Dryer Outlet C
Dryer Outlet 0
Dryer Outlet E

Flash Tank
  NH3 Condenser Outlet
  NH  Condenser Inlet
SCO
SHO
SYO
SOO-A
SOO-B
SOO-C
SDO-D
SOO-E
SFO
SFI
 12600
    97
   323
   802
   223
    75
    23
    16
224000
206000
 18.23
  2.35
  9.61
 43.98
 16.38
  4.95
  2.12
  1.37
 93.916
(No Emission)
2.69 x 10"3
3.42 x 10"4
1.42 x 10'3
4.45 x 10"3 (Total)
        f3
        r3
         4
6.53 x 10
2.43 x 10'
7.32 x 10'
3.13 x 10~
2.02 x 10'
1.02 x 10*2 (Total)
        ,-4
                                                                    ,-2 d
                                                            1.39 x 10
                                                                                              1.39 x 10"2 (Total)
Blend Tank
  NHj Condenser Inlet*
Steam Stripper
  Inlet
  Outlet
SBI
133000'
                                        (No Emission)
                                        (No Emission)
'Pounds  of emission from outlet per  pounds of product processed.
 Production rate (pounds produced per hour) through the knockout was the total  of the average production rates of the
 white and black master lines.
       Includes the average 3.99 pounds per hour of  liquid solvent during reactor dumping (emitting period) and
 89.92 pounds per hour of gaseous emissions.
 Calculated the production rate as the average rate  of the two days testing the dryer outlets.
'Actual concentration of blend tank headspace that should have processed through NHj condenser.
 SBI concentration given as compound; also SBI-2 run not representative due to sampling problems and not included.
"Average of two runs.
 Results are based on Inconclusive field results.
                                                   2-2

-------
                                         Table  2-1.    Continued'
Sample Point Description
                                    Lab No.
                 Hydrocarbon
                 Emission
                 Concentration
                 (PPM •( Benzene)
                      Hydrocarbon
                      Emission
                      Rate
                      (Lbi/Hr)
                      Hydrocarbon
                      Emission
                      Production
                      Rate
                      (Lbs/Lbs')
EMULSION CRUMB PROCESS

White Line
  Roof Vent Ah
  Roof Vent B
  Roof Vent C
  Roof Vent 0
  Dryer Outlet'
ERO-A
ERO-B
ERO-C
ERO-0
EDO
   36
   35
   64
   34
  342
 7.36
 7.43
10.83
 6.03
21.99
                                         1.36 x 10
1.32 x 10
1.94 x 10
1.18 x 10
4.31 x 10'
         r3
         ,-3
r3
r3
rf
,-3
                                                                                                  10.11 x 10 *  (Total)
Herotine Knockout Outlet
EKO
47000
                                                                                B.I
                   4.5 x 10
                                                                                                           -4  b
Stew Stripper
Inlet
Outlet
Black Master Line
Roof Vent Outlet B
Roof Vent Outlet C
Roof Vent Outlet D
Dryer Outlet A
Dryer Outlet B
Dryer Outlet C
Dryer Outlet D

i

ERO-BMB
ERO-BMC
ERO-BMO
EDO-BMA
EDO-BMB
EDO-BMC
EOO-BMO



32
26
25
140
103
81
151

(No Enission)
(No Emission)

0.76
5.04
6.44
9.00
4.96
5.62
10.66



6.12 x 10-5
4.05 x 10~4
5.19 x 10~4
7.26 x 10~4
4.00 x 10~4
4.69 x 10~4
8.60 x 10~4
3.89 x 10"3 (Total)
"founds of emission  from outlet per pounds of product processed.
 Production rate (pounds produced per hour) through the knockout was the total of the average production rates of the
 white and black master lines.
cTotal includes the  average 3.99 pounds per hour of liquid solvent during reactor dumping (emitting period)
 and 89.92 pounds per  hour of gaseous emissions.
 Calculated the production rate as the average rate of the two days testing the dryer outlets.
'Actual.concentration  of blend tank headspace that should have processed through NH3 condenser.
fSBI concentration given as compound; also SBI-2 run not representative due to sampling problems and not included.
"Average of two runs.  -
 Results are based on  inconclusive field results.
                                                       2-3

-------
production rate during the test period.  The results have been separated
into sections for the Solution and Emulsion processes.
     The sum of the emission rates for the Solution Crumb Process tested
was 193 pounds per hour of total hydrocarbon.  The sum of the emission
rates for the Emulsion Crumb Process tested was 104 pounds per hour for
the black master and white production lines.
     The emission rate results are calculated from the test data provided
by monitoring the Borger facility.  However, the plant was operating at
less than maximum production capacity and the results are indicative
only of this production level.
     The production data for the test period was provided by the Phillips
Petroleum personnel and was utilized for the calculation of the pounds
of hydrocarbon emissions generated per pound of product produced.  The
Emission/Production Rate for the Borger facility during the test period
is calculated at 0.042 pounds of hydrocarbon emissions per pound of
production.  This emission rate is only for the plant as tested and does
not account for non-tested or inoperable sources of emissions.  These
results are only representative of the major emission vents because the
testing of all emission points was considered infeasible.
2.2  SOLUTION CRUMB PROCESS TEST RESULT SUMMARY
     Table 2-2 presents the results obtained during testing of the
Solution Crumb Process.  The results provide averages of the test run
analyses and the process conditions during testing.  The individual test
run results are provided in Section 3 and should be referred to for the
discussion of the validity of the test results and the relationship of
the results to the process.
     The ammonia condenser outlet (SFO) sample was obtained during an
approximately five-minute period of uncontrolled emissions occuring when
the flash tank system was incapable of removing the condensed solvent as
fast as the condensation occured.  Hence, the pressure build-up in the
Slowdown Process forced liquid as well as gaseous emissions through the
ammonia condenser outlet.  Expressing emissions in terms of pounds of
emissions per reactor dump is more indicative of the emissions from this
site since preliminary velocity calculations estimated that 90% of the
                                 2-4

-------
                                    Table 2-2.   TEST RESULT  SUMMARY OF  THE SOLUTION  CRUMB PROCESS
Oryer Systen*

flue Gas;
TซP..ฐF
Are*, sq. ft.
• Flow Rate, SCFM
•last Rate. (Ibs/hr)
Total Hydrocarbons
Analysis (Dry Basis)"
Methane
Ethan*
Propane
Butadiene
Pentane
Hexano
Cyclohexane
Benzene
Toluene
Ethyl Benzene
Xylene
Styren*
Volus* Percentages
Oxygen
Nitrogen
Water Vapor
TOTAL
SOO-A

144
3.797
5.46 x 103

43.98

-
-

67.1
-
-
565.5
N.A.
4.1
91.0
19.9
21.1

19.50
72.89
6.00
98.48
500-B

186
3.797
8.56 x 103

16.38

. -
.
-
21.8
-
-
144.9
-
N.A.
30.2
6.9
7.6

20.50
76.40
1.42
98.34
soo-c

196
3.797
7.52 x 103

4.95

.
.
-
S.9
-
-
54.8
-
N.A.
8.8
N.A.
3.4

20.60
76.41
1.74
98.76
SOO-D SDO-E SFIf

193 189
3.797 3.797 .022
6.75 x 103 7.80 x 103 42

2.12 1.37 81.91

N.A.
.
566.4
N.A. N.A. 104,413
728.0
43.144
23.6 11.8 59.434
.
N.A. N.A.
N.A. N.A.
N.A. N.A.
N.A. 4.1

20.10 20.85 0.75
75.26 77.70 75.4
2.09 0.22 0.0
97.46 98.77 96.95
Control Systen* Other Processes*
SFOf SSIh

-
.022
41
gt
93.91"

N.A.
.
238.3
124.501
829.6
45.186
59.457
-
-
.
.
•

2.1
73.03
0.0
98.1
SSOh SB1 SCO

200
0.906
152

g 18.23

N.A.
N.A.
18. 8f 19.3
6,278 1,396
.
25.328
23.588ฐ 9,494
- ' - 63.9
29.1
10.5* 1.046
3.6* 318.8
40.7* 291.3

5.95
22.40
70.60
100.22
SHO

103
0.994
2. 22x WJ

2.3S

.
7.6
-
8.1
-
.
49.0
N.A.
.
8.3
2.9
21.4

19.47
72.66
7.00
99.14
SYO

124
1.917
2.88 x 101

9.61

.
.
.
33.2
-
-
244.4
-
1.2
24.0
4.9
11. S

17.95
67.57
13.00
98.55
'Average of 3 runs.
 Values expressed  In ppei of compound.
cAverage of 2 runs.
 Includes 3.99 Ibs/hr condensed liquid solvent.
N.A. - Trace anount present but not Included  In average values.
'dated on 1 run.
'suspect data, see Section 2 for discussion.
'HO mitt I on rate flows were taken.
 Steaa stripper - no Missions, see Section 2 for discussion.

-------
emissions resulted during a short time period of each reactor dump.   Due
to the testing conditions, the results of the efficiency across the
ammonia condenser give analytical results of an inlet concentration  of
21% hydrocarbon and an outlet concentration of 23% hydrocarbon.  The
test conditions and results at the flash tank ammonia condenser are
discussed further in Section 3.
                                *
     During the test, the cement  filling rates and emptying rates were
essentially equal; therefore, very little pressure build-up resulting in
no flow to the ammonia condenser and to the atmosphere.   A modified
sample location (SBI) for the gas concentration, which should have
flowed through the condenser, was obtained from the blend tank headspace.
The sample was taken at the top of the blend tank from an access port;
the gas sample was obtained when a pressure build-up in the blend tank
caused the weighted port cover to open.  The results from this sample
location were the concentration analysis results as flow measurement was
impractical.
     The steam strippers were considered closed systems and no gaseous
emissions resulted from this sample location.   The efficiency of the
steam stripper system was determined by comparing the outlet and inlet
concentrations.  The inlet sample (SSI) was in the cement form taken
directly from the blend tank drain valve (Figure 5-3).   The outlet
sample (SSO) was in the crumb form taken at a production sample valve at
the exit of the stripper.  The crumb was in a water slurry which flows
from the stripper into the crumb tank.  The outlet sample results reflect
only the VOC contained in the crumb and does not include the VOC in  the
water.  The analytical results of the steam stripper samples are presented
in Table 2-3 along with the solvent removal efficiency.
     The gaseous emission from the operation of filling and emptying the
crumb solution tank (SCO) were vented directly into the atmosphere.   The
gaseous emissions and flow rates were monitored at the stack exit of the
tank.  The total hydrocarbon emission rate was 18.23 pounds per hour and
the average analysis is presented in Table 2-2.
 Plant terminology for solution crumb latex at this stage of production.
                                 2-6

-------
         Table 2-3.  SOLUTION PROCESS STEAM STRIPPER RESULTS

Run #
SSI-1
SSO-1
SSI-2
SSO-2
SSI-3
SSO-3
% Solids
14
29
14
25
18
26
g Hexane
g Dry wt SBR
4.06
.00061
4.34
.0054
3.11

Efficiency (%)
99.98

99.88

	 a

     aSample lost in shipment.
     The Dewatering Screen .Process produced a gaseous emission that was
captured by a hood system and vented to the atmosphere through a roof
vent.  Sample location SHO was maintained at this vent.   The gaseous
emission and flow rate from this roof vent were monitored during a
normal plant operation period.  The average tests results are presented
for sample location SHO in Table 2-2 with a calculated emission rate of
2.35 pounds of hydrocarbon per hour.  The gaseous emission from the
cyclone were vented to the atmosphere through a roof vent (SYO).  Sample
location SYO was maintained at this vent.  The gaseous emission and flow
rate from this roof vent were monitored during a normal  plant operation
period.  The average run results are presented in Table 2-2 with a
calculated emission rate of 9.61 pounds of hydrocarbon per hour.
     A conveyor transported the crumb along the dryer and a system of
five vents emitted the gas emissions to. the atmosphere through roof
vents.  The gaseous emission rates and flow rates from these roof vents
were monitored during a normal plant operation period.  The average test
results are presented in Table 2-2.
                                 2-7

-------
2.3  EMULSION CRUMB PROCESS TEST RESULT SUMMARY
     Table 2-4 presents the test results obtained during testing of the
Emulsion Crumb Process.  The results provide the average of the test run
analyses and the process condition during testing.  The individual test
run results are provided in Section 3 and should be reviewed for the
discussion of the validity of the test results and the relationship of
the results to the process.  Section 4 provides the process information
that should be understood to interpret the sample location results.
     The emulsion process (Figure 2-1) began with a latex being formulated
in the reactors.  The reactor batches are combined in blowdown tanks.
The components in this formulae were highly volatile and were processed
through a flash tank system.  The gaseous emissions from the flash tank
were emitted to the atmosphere through a system of control devices.  The
final control device before emission to the atmosphere was a kerosene
knockout tank.  A sample location (EKO) was maintained here for the
gaseous emissions from the flash tank and steam stripper gaseous exhausts.
Three samples were taken at the kerosene knockout tank exhaust during
the period of flow.  The kerosene knockout exhaust samples were taken by
EPA Method 110 and EPA Method 25.  The samples were analyzed for Total
Gaseous Non-Methane Organics (TGNMO), total hydrocarbons by FID, and for
hydrocarbon compound identification and quantification by GC/FID.   The
results of the TGNMO and the FID total hydrocarbon analysis are presented
in Section 3.  The GC/FID analysis was not successful due to the large
number of organic compounds present in the kerosene.  The chromatogram
and results of this analysis are contained in Appendix B.  The TGNMO
analysis was performed by Pollution Control Sciences of Dayton, Ohio,
and a copy of their results is included in Appendix B.  The flows were
monitored by a flange tap orifice and recorded on a square root chart by
the plant personnel.  The emission rate of the outlet gas stream was
analyzed by an FID method and calculated to be 8.15 pounds of hydrocarbon
per hour.  The rate of the kerosine knockout outlet gas stream as analyzed
by the total gaseous non-methane organic method was calculated to be
2.68 pounds of hydrocarbon per hour.
                                 2-8

-------
                                           Table  2-4.   TEST RESULT SUMMARY  OF  THE  EMULSION  CRUMB PROCESS
ซo
White Line Process*

Flue Gas:
0
. Area, sq. ft.
Flow Rate. SCFN
Mass Rate. (Ibs/hr)
Total Hydrocarbons
Analysis (Dry Baslsjb
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benteno
Toluene
Ethyl Benzene
Xyleno
Styrene
Voluie Percentages
Oxygen
Nitrogen
Water Vapor
TOTAL
ERO-A

93
IS. 2
1.47 x 10*
7.36

-
-
-
N.A.
-
N.A.
-
N.A.
N.A.
-
-
32.0
21.20
78.23
0.00
99.43
ERO-B

99
15.2
1.56 x 104
7.43

-
-
-
N.A.
-
N.A.
-
N.A.
N.A.
-
-
31.0
21.33
78.03
0.00
99.37
ERO-C ERO-0 E00"c

103 98 159
IS. 2 IS. 2 2.4
1.3S x 104 1.43 x 104 5.88 x 103
10.83 6.03 21.99C

.
.
.
N.A.
.
N.A.
-
N.A. N.A. 3.0
N.A. • .9
2.7
.
56. 5 30.1 292.3
21.13 21.13 19.74
77.93 78.03 73.21
0.00 0.00 6.20
99.07 99.17 99.18
ERO-BHB ERO-BMC

108 103
5.4 10.9
1.62 x 103 1.34 x 104
.76 5.04

-
-
-
N.A. N.A.
-
N.A. N.A.
8.0 3.5
-
N.A.
2.3 N.A.
N.A.
18.4 20.1
20.40 20.80
76.80 77.40
0.00 0.00
97.20 98.10
Black Haster
ERO-BMO

108
10.9
1.73 x 104 6.
6.44

-
-
-
N.A.
-
N.A.
3.0
-
N.A.
N.A.
-
19.5
20.90
77.80
0.00
98.60
line Process*
EDO-BMA EOO-BMB

153 129
7.8 5.4
79 x 103 4.07 x 10
9.00 4.96

-
-
-
N.A.
-
N.A.
N.A. N.A.
-
N.A. 2.2
1.4 N.A.
N.A.
122.2 88.3
18.73 19.03
70.51 70.72
8.47 7.92
97.72 97.68
Other Processes*
EOO-BMC EOO-8ND

107 139
3.3 5.5
5.54 x 103 5.90 x 101
5.83 10.66

-
-
-
N.A. N.A.
-
1.8 N.A.
2.1 1.8
-
N.A. N.A.
N.A. 1.5
N.A.
68.1 129.4
19.95 18.86
73.77 69.98
3.48 8.32
97.21 97.17
EKO ฃ50

65
0.049 -
1.33 X 10* •
8.15 -

172.228.7 -
-
-
31.425.6
-
130.4 •
-
-
-
-
-
117.1
13.4
65.2
0.0
99.0
ESI


•
-

-
-
-
-
-
-
-
-
-
-
-
•

-
-
™
        'Average of 3 runs.
        nfalues expressed In ppa of compound.
        cAverage of 2 runs.
        "Samples "ere diluted to within working range of CC/FID.  Values are calculated concentrations based on the dilution ratto.
        N.A. - Trace anounts present but not Included In average values.

-------
7*

R
I
A
C
T
0
R

R
E
A
C
T
0
R
•
R
E
A
C
T
0
R

ป
-ป
P/
••
B
L
0
U
0
0
H
N

B
L
0
U
D
0
U
N

B
L
0
W
D
0
H
N


^•M
-ป
EKO
VACCUM
RECOVERY
TANK




t
PRESSURE
RECOVERY
TANK
f
FLASH
TAHV .
IMK




*



I
KEROSINE
KNOCKOUT
f
KEROSINE
ABSORBER
f
CHILLER
i

STRIPPERL^ **ฅฃ**

•> >"•• inn*



J-ERO J-ERO JxRO J-


WATER
SEPERATOR
'
••^H
'
J - SOAP
CONVERSION
TANK
^J

'
COAGULATION
TANK
I
Bit




1

'NO
INK

IRT A (MANUFACTURE) USED FOR BOTH LINES
PAI


ERO


RESLURRT
•

t
PRESSER
VfKT f
t ฐ
V R
EDO I
R
1

VFWT f
f ฐ
V R
EDO I
IfENT f
^
EDO
1
PRESSURE
AND
PACKAGING

'1
ROOF
•
VENT
, t*
J EDO


ADD CARBON BLACK
FOR BLACK LINE
IT B (PROCESSING)
                                      Figure 2-1.   Emulsion Crumb Process

-------
     The latex formulae front the flash tank was processed through one of
the three steam strippers in the Emulsion Process.   Two of the strippers
("A" and "B") were identical in construction; therefore, only one of
this kind was tested.   The plant designations for the two steam strippers
tested were "B" and "E".   The samples were obtained at the inlet from a
relief valve before the stripper and the outlet sample from a production
sample valve.  The latex was sampled, prepared, and analyzed according
to the EPA Draft Method,  Determination of Residual  Styrene in Styrene
Butadiene Latex by Gas Chromatographic Results (Appendix F).   The con-
centration and efficiency results are contained in Table 2-5 with results
reported as ppm styrene on a dry basis.   The complete results are discussed
in Section 3.
                Table 2-5.   STYRENE REMOVAL EFFICIENCY OF
                     "B" AND "E" STRIPPER COLUMNS IN
                         EMULSION CRUMB PROCESS

Run I.D.
"B" stripper
B-l
B-2
B-3
Average
"E" stripper
E-l
E-2
E-3
Averaqe
Inlet
Cone, (ppm)
Styrene As
Compound
22.5 x 104
15.9 x 104
62.9 x 104
14.4 x 104
27.7 x 104
17.6,x 104
Outlet
Cone, (ppm)
Styrene As
Compound
1.03 x 104
1.30 x 104
1.42 x 104
4
.96 x 10
2.12 x 104
.64 x 104
% Efficiency
95.4
91.8
97.7
95.0
93.3
92.3
96.4
94.0
                                 2-11

-------
     The latex from the stripper systems were stored 1n blend tanks
until required for process finishing.  Two separate finishing production
lines are located at the Borger facility.  The plant designates a "white
line" and a "black master line" for this process finishing system.   The
latex solutions from both lines are treated the same except the black
master had the addition of chemicals for finishing.  The solution is
processed through a series of tanks (Coagulation, Soap Conversion,  and
Reslurry) and the fugitive emission from this processing was captured by
roof vents and emitted to the atmosphere (ERO).  The gaseous emission
and flow rate from the operation of the finishing system of the emulsion
crumb lines were monitored and the results presented in Table 2-4.
     Three samples were taken at each operating roof vent for both  the
white and the black line.  There were four roof vents (A, B, C, D)  over
the white line process and three roof vents operating over the black
line (see Figures 5-10 through 5-13 for vent location with respect  to
process).  A modified EPA Method 110 was followed in order to obtain the
samples.  The flows were measured by a calibrated vane anemometer.   The
samples were analyzed by GC/FID.  The results for the white line vents
are presented in Tables 3-12 to 3-15 and the black line in Tables 3-17
to 3-19.  The white line produced significantly higher results than the
black line.   No explanation for these results is readily apparent.
Sample point ERO-BMB shows consistent concentration results but lower
emission levels.   This was due to an inoperable fan at this location.
The gas flow resulted from convection currents.  The ERO-BMA Vent had no
gas flow at all and therefore no sample was collected.
     The latex was processed from the finishing tank through a presser
to remove water.   This crumb form is conveyed through a dryer system.
The white line has two dryer systems, and the black master has three
dryer systems.  But only the dryers operating during testing were sampled.
     Three samples were taken at the operable dryer vents on both the
white and black line.   The white line was operating at a low level, and
only one dryer was utilized.  Of the two vents (EDO-A, EDO-B) on the
operating dryer,  only one (EDO-A) was functioning at the time of the
test.  Similarly, on the black line only one dryer was utilized, as the
other was down for maintenance.  All four associated dryer vents were
                                 2-12

-------
functioning.  The samples were obtained utilizing a modified EPA Method 110
and analyzed by GC/FID.  Flows were measured by a calibrated vane anemo-
meter on the black line and by EPA Methods 1-4 on the white line.  A
pseudostack was erected on the white line vent in order to conform to
the distance requirements of EPA Methods 1 and 2.  The white line results
and the black line results are listed in Table 2-4.
                                 2-13

-------
                       3.0  DISCUSSION OF RESULTS

     The two production processes, Solution Crumb Process and Emulsion
Crumb Process, are discussed separately and include the following at
each sampling location how the samples were taken, how the results were
obtained, the analytical results, and any deviations from the standard
and draft methods.
3.1  SOLUTION CRUMB PROCESS
3.1.1  Solution Crumb Process;  Flash Tank Ammonia Condenser
     The results of the hydrocarbon analysis show high levels of hydrocarbon
emissions at the outlet (Table 3-2) with a total  gaseous hydrocarbon
emission average of 89.92 Ib/reactor dump.   An average of 3.97 pounds/
reactor dump of liquid was also collected.   The sample was collected
over the 5-minute uncontrolled emission flow.   There were three samples
taken at the outlet.  The emissions from each run are fairly consistent
and the average is a good representation of the emission rate.   The
inlet sample collected at the same time shows less total hydrocarbons
going in (81.91 Ib/reactor dump) than are going out.  The diagram of the
flash tank accumulator and the ammonia condenser shows the inlet of the
accumulator to be at the top of where if a large flow of gases and
liquids are released due to a reactor dump, it overloads the accumulator
and both are expelled through the condenser.  Due to the modified sampling
point at the inlet, no liquid emissions were found at the inlet, but a
substantial amount was expelled from the outlet.   The high flow during
this period rendered the ammonia condenser ineffective and the liquid
solvent was emitted without a temperature drop.
     During the blowdown cycle, the ammonia condenser vent remained open
for a longer period than expected (approximately 15 minutes), the flow
rate was higher than expected, and both gas and condensate were emitted.

-------
                   Table  3-1.   SUB-SUMMARY OF RESULTS  AT  THE  SOLUTION
                                CRUMB PROCESS -  FLASH TANK INLET
Ttซt tun 9
Date
Flue 6as Tea? (ฐF)
Hut Art* (Ft2)
Flu* Flow Rite (SCFH)
Total Hydrocarbon
Eป1ss1on Rate (U)/Hr)
Analytical Results* (PPM)
Methane
Ethan*
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (at Capd)
Oxygen
Nitrogen
Hater Vapor
Total Percentage
SFI-1 SFI-2 SFI-3 Average of Runs
5/31/80 5/31/80 5/31/80
— • ซj.
0.022 0.022 0.022
38 46 41 42
64.18 73.53 88.01 81.91
As Benzene As topd As Benzene As Capd As Benzene As Capd As Benzene As Capd
2.4 10.3 3.2 13.6 — — N.A. M.A.
— — — — —
147.2 165.1 810.9 1261.3 162.6 252.9 373.6 566.4
98067.5 123387.0 55819.7 70231.4 95074.8 119622.0 82987.4 104413.5
778.3 790.5 481.9 487.7 894.8 905.7 718.3 728.0
57382.6 48220.7 33418.1 28082.4 63223.4 53129.0 51341.4 43144.0
64308.0 54040.3 68915.0 57911.8 78956.3 66350.0 70726.4 59434.0
—
—
—
I
... ... ... ... ... ... ... ...
220686 226633 159448 157988 238312 240259 206147 208286
22.58 15.80 24.02 20.80
0.77 0.95 0.52 0.75
72.3 78.4 75.5 75.40
0.0 0.0 0.0 0.0
95.65 95.15 100.04 96.95
*No temperatures were recorded for these tests as there was no location available.

 Gaseous results only.
N.A. - Trace aaount's present 1n results but not average.
                                              3-2

-------
                      Table 3-2.   SUB-SUMMARY  OF  RESULTS  AT  THE  SOLUTION
                                CRUMB  PROCESS -  FLASH TANK  OUTLET
Test tun t
Date
flut Gas Ttap (ฐF)
Flue Area (Ft2)
Flue Flow Rite (SCFH)
Total Hydrocarbon
Emission Rate (Lb/Hr)
SFO-1
5/31/80
64
0.022
19
81.26
SFO-2
5/31/BO
68
0.022
45
95.42
SFO-3
5/31/80
68
0.022
41
93.08
Average of Runt
66
0.022
41
89.92
{Analytic*! Results1 (PPM)   As Benzene   As Capd  'As Benzene   As Capd  As Benzene    As Cซpd  As Benzene   As Cซpd
Methane
Cthane
Propane'
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
146.0
...
149.8
101844.0
768.5
40412.0
72357.0
—
—
...

*••
215676
. 623.4
...
233.0
128138.0
777.8
33960.0
60804.2
—
—
—
..V
i
...
224536
8.2
...
147.2
99940.0
795.5
57680.0
65534.0
—
—
—
...
••*
224106
35.0
—
228.9
125743.0
805.2
48471.0
55071.0
—
—
—
—
•ป•ป
230354
...
...
162.6
95075.0
694.8
63223.0
74321.0
—
—
—
...
• •*
233726
...
—
252.9
119622.0
905.7
53129.0
62497.0
—
—
...
—
•••
236406
N.A.
—
153.2
98953.0
819.6
53772.0
70754.0
—
—
—

ป"~
224452
N.A.
—
236.3
124501.0
829.6
45186.0
59457.0
...
...
—
—
...
230212
 Volume Percentages (X)
Hydrocarbon (as Cupd)
Oxygen
Nitrogen
Water Vapor
22.45
2.0
73.2
0.0
23.04
2.6
74.0
0.0
23.64
1.6
71.9
0.0
23.04
2.1
73.0
0.0
 Total Percentage
                                97.65
99.64
                                                                        97.14
                                        98.1
 'Caseous results only.
 N.A.  - Trace amounts present In results but not averaged.
                                                    3-3

-------
The sampling method was modified to allow accurate flow measurements
with an annubar (Section 5.1.1).  A condensate trap was placed in line
so that the liquid emissions could be collected and analyzed separately.
     The analytical procedure was changed in order to include the
condensate and to accomodate the high hydrocarbon levels.  The condensate
was diluted and analyzed by direct injection into the GC/FID.  The
analytical results in Table 2-1   reflect the addition of the condensate
analysis.  The levels of hydrocarbons in the sample bag required dilution
to be within the working range of the instrument.  Corrections of the
concentrations for the dilution were made in the field results (Appendix B)
and are reflected in the grand summary (Table 2-1).
3.1.2  Solution Crumb Process:  Blend Tank
     The results listed are of the headspace above the tank which
periodically vented through a pressure relief cap on the side of the
tank.  There appeared to be no set time for this release and no emission
data can be assumed due to the periodic nature of the emissions.
3.1.3  Solution Crumb Process:  Steam Stripper
     No gaseous emissions are vented from the steam stripper; however,
samples of the latex cement (inlet) and of the latex crumb (outlet) were
taken.  The process was operating normally and the samples were collected
and analyzed as described in the EPA Draft Method for the Determination
of Residual Cyclohexane (Appendix F).  The results are shown in Table 2-3.
A complete description of the sampling and analytical methods is found
in Section 6.1.3.   Briefly, the samples were collected by an integrated
latex grab method and the samples were analyzed by GC/FID.
3.1.4  Solution Crumb Process:  Crumb Tank
     The crumb tank has a stack which is open to the atmosphere and gas
samples of the emissions above the crumb tank were collected there
(Figure 5-3).   The process was operating normally during the course of
the sampling.   Flow measurements were taken from anemometer readings.
The temperature was monitored using a type-K thermocouple wire and
digital display.   Samples were collected following the modified EPA
Method 110 (Appendix G).  Analysis was performed by GC/FID (Appendix F).
The results (Table 2-2) show that an average of 18.231bs/hr of hydrocarbons
are emitted during normal process operation.
                                 3-4

-------
     In tests 1, 2 and 3, the concentration of total  hydrocarbons were
11,155 ppm, 10,375 ppm and 16,493 ppm, respectively (Table 3-3).
3.1.5  Solution Crumb Process:  Dewatering Screens
     A hood over the dewatering screens draws the emissions up to a vent
on the roof.  Gas samples were collected at the vent according to the
modified EPA Method 110 (Appendix G).  In order to obtain accurate flow
measurements, a pseudo-stack was attached to the existing vent to conform
to the distance requirement of EPA Methods 1 and 2 (Section 5.1.5).  The
pseudo-stack had no effect on the analytical results, but allowed for
accurate flow measurements.  Analysis of the gas samples was performed
by GC/FID (Appendix F).  The process was operating normally and the
results (Table 2-3) can be considered normal.  The average emission rate
of hydrocarbons during dewatering screen operation was 2.35 Ibs/hr.
Table 3-4 identifies each compound found in the sample for each run and
reports the process data monitored during each run.  The gas sample bag
in the second run was virtually empty.  The concentration results obtained
are considerably lower:  test 1 had 110 ppm total hydrocarbon; test 2
                          i
had 36.3 ppm total hydrocarbon; and test 3 had 88.7 ppm total hydrocarbon.
The results from the second test were not considered valid and, therefore,
were not included in the averages of the dewatering screen emission.
3.1.6  Solution Crumb Process:  Cyclone Vent
     The average emission for the three runs at the cyclone vent was
9.61 Ibs/hr of hydrocarbons.  The average appears to be a good represen-
tation of the expected emissions; the flow rate remained constant over
three tests and the total concentration of hydrocarbons remained around
300 ppm for each test (Table 2-2).
     Gas samples were collected at the outlet of the cyclone vent following
the modified EPA Method 110 (Appendix G).  In order to obtain accurate
flow measurements, a pseudo-stack was attached to the existing vent to
conform to the distance requirement of EPA Methods 1 and 2 (Section 5.1.5).
The pseudo-stack had no effect on the analytical results, but allowed
for accurate flow measurements.  Analysis of the gas samples was performed
by GC/FID (Appendix F).  The process was operating normally and the
results can be considered normal.  The sub-summaries (Table 3-5) identifies
each compound found in the sample for each run and reports the process
data monitored during each run.
                                 3-5

-------
                   Table 3-3.   SUB-SUMMARY  OF  RESULTS AT THE SOLUTION
                               CRUMB PROCESS - CRUMB TANK  VENT
Tttt lun 1
Date
Flu* Gas Teop (ฐF)
Flue Arti (Ft2)
Flu* Flew Rate (SCFH)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Result! (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Capd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
SCO-1
6/10/80
194
0.906
132
13.91
As Benzene As Cepd
0.9 3.9
13.6 31.3
7.8 12.2
730.9 958.6*
—
—
8156.7 8156.7
70.7 70.7
24.6 20.0
134B.5 926,4
463.6 311.5
344.1 301.3*
11521.6 11154.6
1.12
6.03
22.61
70.60
100.36
SCO-2 SCO-3
6/10/80 6/10/80
204 200
0.906 0.906
157 168
15.17 25.62
As Benzene As Capd As Benzene As Capd
... ... ... ...
— 4.9 12.0
8.1 12.6 19.6 33.0
741.0 971.6* 1593.6 2256.4*
—
—
8090.9 8090.9 11000.0 11875.0
47.6 47.6 73.4 73.4
28.5 23.2 50.5 44.3
1160.3 798.8 1898.6 1411.0
357.2 239.9 . 558.7 405.1
216.6 189.8* 404.9 382.9*
10650.3 10374.6 15604.3 16493.1
1.04 1.65
5.86 5.94
22.40 22.20
70.60 70.60
99.92 100.39
Average of Runs
199.7
0.906
152
16.23
As Benzene As dipt)
N.A. N.A.
N.A. N.A.
11.8 19.3
1021.8 1395.6
—
—
9202.5 9494.2
63.9 63.9
34.5 29.1
1469.2 1046.1
459.9 318. 8
321.9 291.3
12585.5 12658.3
1.27
5.95
22.40
70.60
100.22
*Value 1s the sun of one or more unknown compounds closely associated with the Identified compound.  Individual
 concentrations for each compound are Included In the computer data sheets, Appendix A.

N.A. • Trace amount* present In results but not averaged.
                                                   3-6

-------
Table 3-4.   SUB-SUMMARY  OF  RESULTS  AT THE  SOLUTION
       CRUMB  PROCESS -  DEWATERINP  SCREEN  HOODS
Ttit lun •
Date
Flut as TMP (ฐF)
Flut Art* (Ft2)
Flue Flow Rate (SCFH)
Total Hydrocarbon
Eaittlon Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Mexane
Cyclohexane
Benzene
Toluene
Cthylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentage* (X)
Hydrocarbon (at Cupd)
Oxygen
Nitrogen
Hater Vapor
Total Percentage
SHO-1 SW-2* SHO-3 Average of Runt*
6/10/80 6/10/80 6/10/60
101 104 104 103
0.994 0.994 0.994 0.994
2.19 x 10S 2.24 x 103 2.25 x 103 2.22 x 103
2.73 0.96 1.98 2.35
As Benzene At Capd At Benzene At Capd At Benzene At Capd At Benzene At Cupd
... ...
3.3 7.4 — — 3.4 7.6 3.3 7.6
3.5 4.5b 1.5 1.9b 9.0 11.6b 6.2 8.1
—
— — — —
43.7 43.7 20.6 20.6 54.3 54.3 49.0 49.0
— 	 	 — 3.0 3.0 N.A. N.A.
...
13.8 9.5 6.9 4.7 10.3 7.1 12.1 6.3
5.1 3.4 2.4 1.6 3.5 2.4 4.3 2.9
46.2 40.5 6.5 7.5 2.7 2.3 24.5 21.4
115.6 110.0 39.9 36.3 66.2 68.7 97.4 96.3
0.01 0.00 0.01 0.01
19.62 19.16 19.62 19.62
73.19 71.24 73.56 72.37
7.00 7.00 7.00 7.00
99.62 97.40 100.19 99.14
,*6ag for SHO-2 empty - results unacceptable for averaging only SHO-1 and SHO-3 In average remits.

bVa1ue  It the tun of one or ซore unknown compounds closely associated with the Identified compound.
 concentrations for each compound are Included in the computer data sheets, Appendix A.

N.A. •  Trace amounts present In results but not averaged.
                                                                   Individual
                                 3-7

-------
                 Table 3-5.   SUB-SUMMARY OF  RESULTS AT THE SOLUTION
                              CRUMB PROCESS -  CYCLONE VENT
Test Run f SYO-1 SYO-2 SYO-3
Date 6/5/80 6/5/80 6/10/60
Flue CM Ttap (ฐF) 125 124 124
Flue Are* (Ft2) 1.917 1.917 1.917
Flue Flow Rate (SCFH) 3.08 x 103 2.76 x 103 2.76 x 103
Total Hydrocarbon
Emission Rate (Lb/Hr) 12.12 9.59 7.12
Analytical Results (PPM) As Benzene At Capd As Benzene As Capd As Benzene As Cnpd
Methane
Ethane — - — — — — — •-
Propane
Butadiene 27.9 37.1a 26.2 34.9* 21.0 27. 6a
Pentane — — — ••• — ---
Hexane — — — — — —
Cyclohexane 292.0 296.1 254.1 257.8 179.3 179.3
Benzene
Toluene 1.7 1.4 1.1 0.9 1.7 1.4
Ethylbenzene 41.4 29.0 29.0 20.2 33.2 22.6
Xylene 8.3 5.7 5.3 3.6 6.0 5.4
Styrene 12.9 11.4 18.7 16.6 7.6 6.6
Total Hydrocarbon 384.3 380.7 334.0 334.0 250.8 243.1
Volume Percentages (X)
Hydrocarbon (as Cnpd) 0.04 0.03 0.02
Oxygen 17.92 17.92 18.01
Nitrogen 67.86 67.66 66.99
Water Vapor 13.00 13.00 13.00
Total Percentage 98.82 98.61 98.02
'value Is the sum of one or more unknown compounds closely associated with the Identified compound.
Average of Runs
124
2.88 x 103
9.61
As Benzene As Cmpd
... —
—
25.0 33.2
—
—
241.8 244.4
—
1.5 1.2
34.6 24.0
7.2 4.9
13.0 11.5
323.1 319.2
0.03
17.95
67.57
13.00
98.55
Individual
concentrations for each compound are Included In the computer data sheets, Appendix A.
                                           3-8

-------
3.1.7  Solution Crumb:  Dryer Vents
     There were five vents carrying emissions off the dryer (Section 5.1.6).
Flow and temperature measurements were made during the test at each
dryer vent (Section 6.1.7).  The vents were modified with sample ports
for the flow measurements.  An average of the measurements for each run
appear in the sub-summaries (Tables 3-6 to 3-10).  Gas samples were
collected from each dryer vent following the modified EPA Method 110
(Appendix G).  Analysis was performed by GC/FID (Appendix F).   The
results (Table 2-2) show decreasing emission rates as the sample location
progress down the length of the dryer.  The hydrocarbon emission rate
pattern decreased from 87.78 Ib/hr at Vent A to 2.73 Ib/hr at  Vent E.
This was the expected pattern.  Test run 3 showed considerably higher
concentration emissions at every dryer vent.  The flow rates were constant
over all 3 tests.   The higher concentrations observed during test run 3
may have been caused by a higher temperature in the dryer during the
third run or by decreased stripper efficiency during this period.   Since
it was impractical to measure all points simultaneously, the resolution
of this problem is uncertain.   It is quite possible, therefore, that the
calculated average is not representative of the actual emission rate.
3.2  EMULSION CRUMB PROCESS (Figure 2-2)
     Before reading the discussion of results of the emulsion  crumb
process it would be advisable to become familiar with the process
description (Section 4.2).  Briefly, this is because the process has two
products.   Both products are processed the same in the "manufacturing"
system (Sections 3.2.1, 3.2.2).   The products differentiate into 1 white
line or 1 of 3 black master lines in the process finishing system.   The
processes differ only in additives in the latex.  The results  of the
white line process are discussed in Sections 3.2.3 and 3.2.4.   the
results of the black master process are discussed in Sections  3.2.5 and
3.2.6.
3.2.1  Emulsion Crumb Process:  Kerosene Knockout
     The kerosene knockout controls the volatile emissions resulting
from the manufacturing system (Figure 2-2).  The latex is formed in the
reactors from sytrene and butadiene.  Upon completion of the reaction,
                                 3-9

-------
                Table 3-6.   SUB-SUMMARY  OF  RESULTS AT THE SOLUTION
                               CRUMB  PROCESS -  DRYER  VENT A
Tซtt Run f
Date
Flue G*( IMP (ฐF)
Flue Area (Ft2)
Flue Flow Rite (SCFM)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Nexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Capd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
SDO-A-1 SOO-A-2 SOO-A-3 Average of Runs
6/5/80 6/5/80 6/10/80
131 132 169 144
3.797 3.797 3.797 3.797
6.38 x 103 4.64 x 103 5.36 x 103 5.46 x 10 3
40.23 22.65 69.06 43.98
As Benzene As Capd As Benzene As Capd As Benzene As Capd As 'Benzene As Cnpd
... ... — — — —
—
28.0 37.3* 22.3 29.7* 102.4 134.38 50.9 67.1
—
—
421.7 427.7 344.6 349.5 919.3 919.3 561.9 565.5
2.8 2.8 N.A. N.A.
4.3 3.3 2.5 2.1 8.2 6.7 5.0 4.1
119.4 83.3 84.3 58.8 190.1 130.8 131.2 91.0
24.9 17.0 18.0 12.3 45.5 30.6 29.5 19.9
23.6 21.0* 17.5 15.6* 30.7 26. 98 24.0 21.1
621.9 589.8 489.2 468.0 1299.0 1251.4 802.5 766.7
0.06 0.05 0.12 0.08
19.27 19.74 19.51 19.50
72.47 73.70 72.51 72.89
6.00 6.00 6.00 6.00
97.80 99.49 98.14 98.48
'value Is the sum of one or more unknown components closely associated with the Identified compound.  Individual
 concentrations  for each compound are Included In the computer data sheets, Appendix A.
N.A. - Trace amounts present In results but not averaged.
                                                  3-10

-------
                 Table 3-7.   SUB-SUMMARY OF RESULTS AT THE  SOLUTION
                                CRUMB  PROCESS - DRYER  VENT B
Ttst Run I
Date
Hut Cซi TMP (ฐf)
riut Area (Ft2)
Flu* Flow Rate (SCFM)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Result* (PPM)
Methane
Ethane
Propane
Butadiene
Pantane
Hexane
Cyclohexane
Benzene
Telutn*
Ethylbenzene
kylene
Styrtnc
Tottl Hydrocarbon
Voluw Percentages (X)
Hydrocarbon (at Capd)
Oxygen
Nitrogen
Hater Vapor
Total Percentage
$00-8-1 $00-8-2 SOO-B-3 Average of Rum
C/S/BO 4/5/80 6/10/80
182 104 193 186
3.797 3.797 . 3.797 3.797
9.54 x 103 8.60 x 103 7.35 x 103 8.56 x 1C3
14.96 4.99 29.20 16.38
As Benzene As Capd As Benzene As Capd As Benzene As Capd As Benzene As Capd
... _. — — — _.
_. —
5.1 6.8* 1.7 2.3* 42.9 56.3 16.6 21.8
— — — — — —
— — .— — — . —
92.5 93.5 28.4 28.6 312.4 312.4 144.5 144.9
... — _. ... ... .—
1.2 1.0 — — 2.0 1.6 N.A. N.A.
51.2 35.7 19.1 13.3 60.4 41.6 43.6 30.2
11.9 8.1 4.7 3.2 13.8 9.3 10.2 6.9
9.5 6.5 7.9 7.0 8.4 7.3 8.6 7.6
171.4 153.6 61.6 54.6 439.9 428.5 223.5 211.4
•
0.02 0.00 0.04 0.02
20.50 20.50 20.50 20.50
76.30 76.30 76.60 76.40
1.42 1.42 1.42 1.42
98.24 98.22 98.56 9B.34
'*Value Is the sun of one or axire unknown compounds closely associated with the Identified compound.   Individual
 concentrations for each compound are Included in the cooput.tr data  sheets. Appendix A.

 N.A. - Trace anounts'present 1n results but not averaged.
                                                    3-11

-------
              Table 3-8.   SUB-SUMMARY OF  RESULTS AT THE SOLUTION
                           CRUMB PROCESS -  DRYER VENT C
Test Run f
Date
Flue CM Tซap (ฐF)
Flue Area (ft2)
Flue Flow Rate (SCFH)
Total Hydrocarbon
Emission Rate (Lb/Nr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrogen (as Cupd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
SOO-C-1 SOO-C-2 SDO-C-3 Average of Runs
6/5/80 6/5/80 6/10/80
196 197 194 196
3.797 3.797 3.797 3.797
8.57 x 103 7.28 x 103 6.73 x 103 7.52 x 103
1.34 0.92 12.60 4.95
As Benzene As Copd As Benzene As Cซpd As Benzene As Cmpd As Benzene As Cnpd
... ... ... ... . — ... . — ...
—
—
0.7 1.0 0.2 0.3 12.5 16.4 4.5 5.9
—
—
9.6 9.7 8.2 8.4 146.4 146.4 54.7 54.8
—
1.1 0.9 N.A. N.A.
2.7 1/9 1.7 1.2 34.0 23.4 12.8 8.8
7.9 5.3 N.A. N.A.
4.5 4.0 3.4 3.0 3.7 3.3 3.9 3.4
17.5 16.6 13.5 12.9 205.6 195.7 75.9 72.9
t
0.00 0.00 0.02 0.01
20.83 20.83 20.14 20.60
76.74 76.74 75.76 76.41
1.74 1.74 1.74 1.74
99.31 99.31 97.66 98.76
N.A. - Trace amounts present in results but not averaged.
                                           3-12

-------
                 Table 3-9.   SUB-SUMMARY  OF  RESULTS AT THE SOLUTION
                                CRUMB  PROCESS -  DRYER  VENT D
Tซt tun ป
Oat*
Flue Gas Teปp (ฐF)
Flut Area (Ft2)
Flu* Flow Rite (SCFH)
ToUl Hydrocarbon
Ealtsion Rate (Lb/Hr)
Analytical Results (PPM)
Hethene
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Tolutnt
Ethyl benzene
Xyline
Styrene
Total Hydrocarbon
Volume Pcrc*ntag*( (X)
Hydrocarbon (as C*pd)
Oxygen
Nitrogen
Hater Vapor
"..Total Percentage
SOO-D-1 SOO-0-2 $00-0-3 Average of Runt
t/5/80 t/S/BO 6/10/80
189 192 197 193
3.797 1.797 3.797 3.797
7.54 x 10J 6.86 x 103 5.85 x 103 6. 75 x 103

1.16 0.29 4.90 2.12
As Benzene As Capd As Benzene As Ce$d Ai Benzene As Cซpd As Benzene As Ciipd
... ... ... — ...
—
—
0.1 0.1 — — 4.9* 6.5 H.A. N.A.
—
—
4.4 4.4 4.5 4.5 61.9 61.9 23.6 23.6
—
— — — 0.4 0.3 N.A. N.A.
0.7 0.5 — — 16.5 11.3 N.A. N.A.
3.7 2.5 N.A. N.A.
10.1 9.0 — — 4.5 3.9 N.A. N.A.
15.3 14.0 4.5 4.5 91.9 86.4 23.6 23.6

0.00 0.00 0.01 0.00
20.17 20.17 19.97 20.10
75.39 75.39 75.00 75.26
2.09 2.09 2.09 2.09
97.65 97.65 97.07 97.46
*Value is the sun of one or mart unknown components closely associated with the Identified compound.  Individual
 concentrations for each compound are Included In the computer data sheets. Appendix A.

N.A. - Trace aaounts present In restuls but not averaged.
                                                 3-13

-------
                    Table  3-10.   SUB-SUMMARY  OF  RESULTS AT THE SOLUTION
                                    CRUMB  PROCESS -  DRYER  VENT E
SOO-E-l SOO-E-2 SDO-E-3 Avtrtge of Runt
Date
Flue Gas Temp (ฐF)
Flue Area (Ft2)
Flue Flow Rate (SCFM)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon as (Cnpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
6/5/80 6/5/80 6/10/60
168 187 194 189
3.797 3.797 3.797 3.797
8.74 x 103 7.93 x 103 6.74 x 103 7.60 x 103

0.23 0.83 3.04 1.37
As Benzene As Copd As Benzene As Cซpd As Benzene As Cnpd As Benzene As Cmpd
—
—
—
2.9 3.9* N.A. N.A.
—
—
0.7 0.7 2.6 2.7 32.1 32.1 11.8 11.8
—
-7- 	 	 1.1 0.9 N.A. N.A.
7.5 5.2 N.A. N.A.
1.7 1.1 N.A. N.A.
1.9 1.7 7.8 6.9 4.2 3.7 4.6 4.1
2.6 2.4 10.4 9.6 49.6 46.9 16.4 15.9

0.00 0.00 0.00 0.00
20.85 20.85 20.85 20.85
77.83 77.83 77.43 77.70
0.22 0.22 0.22 0.22
98.90 98.90 98.50 98.77
*Valve is the sun of one or acre unknown components closely associated with the identified compound.  Individual
 concentrations for each compound are included in the computer data sheets,  Appendix A.
N.A. - Trace amounts present in results but not averaged.
                                                  3-14

-------
the latex flows through the blowdown tanks to the flash tank, which 1s
under vacuum.  In the flash tank, most of the residual butadiene is
flashed off.  The emissions pass through a pressure recovery tank, a
vacuum recovery tank, a brine chiller, a kerosene absorber and a kerosene
knockout (Section 4.2).  Briefly, the pressure recovery tank and the
vacuum recovery tank act as recycling devices.  Hydrocarbons are condensed
out of the gas flow by changing the pressure of the gas.  The brine
chiller serves as a control device by condensing some of the hydrocarbons.
The kerosene absorber further reduces the hydrocarbon concentration of
the emission.  The kerosene knockout removes the liquid kerosene,
Finally, the emissions are vented to the atmosphere.  Also, emissions
from the steam stripper pass through the chiller, kerosene absorber and
kerosene knockout.  Therefore, the samples collected at the outlet of
the kerosene knockout (Figure 5-9) represent the combined emissions from
the flash tank and steam stripper.
     Three gas samples were collected at the kerosene knockout outlet
following the modified EPA .Method 110 and following EPA Method 25.  The
samples were collected during the period of flow as recorded by the
plant.
     Analysis was performed by GC/FID for identification of hydrocarbons
in the sample.  No results are presented from the GC/FID analysis because
the chromatogram becomes unreadable due to the petroleum products asso-
ciated with kerosene.  FID analysis was performed on diluted samples.
The results were expressed as total hydrocarbons as propane since propane
was used to calibrate the FID (Table 2-4).  The dilution ratios were
calculated from GC/TCD analyses (Appendix F).  The results from EKO-1
were based on only one sample injection and therefore are somewhat
unreliable.  The other two runs are the average of two repeatable (within
5%) injections.  Analysis of the EPA Method 25 (TGNMO) samples was
performed by Pollution Control Science, Inc.  The results of the analysis
were compared with the FID results (Table 3-11).  The results from the
two methods were analyzed differently and are not comparible.
     There are no reasons for this; however, it must be pointed out that
TGNMO results are expressed as C^ and the FID results are as propane and
do not take into account the various response factors of differing
hydrocarbons.
                                 3-15

-------
                           Table  3-11.   SUB-SUMMARY  OF  RESULTS AT THE  EMULSION CRUMB
                                       PROCESS -  KEROSENE  KNOCKOUT  ON 6/16/80

Location:


Test Run (Lab •) EKO-1 TGNMO. EKO-2 TGNMO. EKO-3 TGNMO. Average of Run*
EKO-10 EKO-20 EKO-30 EKO-1. 2, 3






CO
I
ป-*
Ol


Flue Gas Te*pฐF 68
Flue Area F2 0.049
Flue Flow Rate SCFH 1.32 x 101
VOC Emission Rate
(Ib/hr) 8.03
Analysis (dry basis) As Benzene As Compound
Methane. PPM ' 28.301.7 167.537.0
Ethane. PPM
Propane, PPM
Butadiene. PPM 14.407.3 31,833.7
Cyclohexane. PPM 231.8 323.0
Benzene, PPM
Toluene, PPM
Ethylbenzene, PPM
Xylene, PPM
Styrene. PPM 208.9 112.2
61 65
0.049 0.049
1.34 x 101 1.33 x 101
6.69 9.74
As Benzene As Compound As Benzene As Compound
22,317.2 132,110.0 36.664.0 217,039.0
—
.-
13,142.5 24,034.3 21.004.5 38,408.8
5.6 7.8 43.3 60.4
--
--
77.8 41.8 367.5 197.3
65
0.049
1.33 x 101
8.15
As Benzene As Compound
29,094.3 172.228.7
..
-.
17,184.0 31,425.6
93.6 130.4
.-
--
218.1 117.1
TOTAL HCa
                     46.149.7   199.805.9   109.251
35,453.1   156.193.9
39,021
58.079.3  255.705.5
52.752
45^590.7   203.901.8
HC. %
Oxygen, %
Nitrogen, X
Water Vapor. %
TOTAL, X
19.98
13.5
65.6
0.0
99.1

15.62
13.3
65.0
0.0
93.9

25.57
13.3
65.1
0.0
104.0

20.39
13.4
65.2
0.0
99.0

*Saซples were diluted to within working range of the GC/FIO.  Values are calculated based on the dilution ratio (Appendix B).
 TGNMO values reported by Pollution Control Science in total ppซ as Cj.

-------
3.2.2  Emulsion Crumb:  Steam Stripper
     The latex flows from the flash tank to one of three steam strippers.
The Phillip's designations are A, B, and E.  Lines A and B are older
(circa 1941) and smaller than line E (circa 1953).  The stripper on line
E 1s approximately twice the size of the strippers on lines A and B.
Latex samples were collected at B and E since A and B were assumed to be
the same.  The gaseous emissions from the steam stripper flow through
the same chiller, kerosene absorber and kerosene knockout, as the emis-
sions from the flash tank.  Therefore, the emissions collected at the
outlet of the kerosene knockout represent the total emissions from both
the flash tank and steam stripper.  Latex samples were collected at the
inlet and outlet of steam strippers B and E following the integrated
grab method (Appendix G).
     After preparation of the samples as described in the EPA Draft
Method for the Determination of Residual Styrene (Appendix F) the samples
were analyzed by GC/FID.  The removal efficiency of residual styrene by
steam stripping was 94.5% (average, Table 2-5).  The process was operating
                          i    .
normally, maximum capacity, when the samples were collected.  Therefore,
the results reflect only on this level of production (Process Data,
Appendix D).
3.2.3  Emulsion Crumb:  White Line Process:  Roof Vents
     After steam stripping the latex flows into a storage tank until it
is required  for processing.  The latex will be processed either in the
white line or in one of the black master lines.  If it is processed in
the white line, it flows through a series of treatment tanks:  the blend
tank, the coagulation tank, the soap conversion tank, the water separator,
the reslurry, and presser (Figure 5-11).  The pelletized crumb is then
dried and packaged.  This entire process is performed inside one building.
Four vents in the roof are spaced along the process (Figure 5-11).  Roof
vent A is located above the coagulation tank, conversion tank and
dewatering screen.  The average emission rate during operation was 7.36
Ib/hr (Table 3-12).  The concentration results from the first two runs
were based on only one sample injection.  All three runs were included
in the average since there is no way to determine which test runs were
the most representative.  Roof vent B is located above the presser.  The
                                 3-17

-------
                  Table  3-12.   SUB-SUMMARY  OF-RESULTS  AT THE EMULSION
                          CRUMB PROCESS  - WHITE LINE ROOF VENT A
Ttst Run ff
Date
Flue Gas Ttop (ฐF)
Flu* Art! (Ft2)
Flu* Flew Rite (SCFM)
Total Hydrocarbon
Enution Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Qgpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
ERO-A-1* ERO-A-2* ERO-A-3 Average of Runs
6/1/80 6/2/80 6/3/80
89 98 93 93
15.155 15.155 15.155 15.155
1.62 x 104 1.4 x 1C4 1.4 x 10* 1.47 x 10*
5.94 3.66 12.47 7.36
As Benzene As Capd As Benzene As Capd As Benzene As Capd As Benzene As Cmpd
...
—
—
0.2 0.3 0.4 0.5 N.A. N.A.
—
0.1 0.2 1.1 1.2 N.A. N.A.
—
2.6 2.6 11.2 11.2 N.A. N.A.
1.4 1.1 — — N.A. N.A.
i
... — — — —
27.4 26.2 17.9 15.7 61.6 54.1 35.7 32.0
27.4 26.2 22.3 19.9 74.5 67.0 35.7 32.0
0.00 0.00 0.01 0.00
21.60 20.90 21.10 21.20
78.10 78.30 78.30 78.23
0.00 0.00 0.00 0.00
99.70 99.20 99.41 99.43
*0n1y one sample Injection was used to calculate concentrations.
N.A. - Trace amounts present In results but not averaged.
                                           3-18

-------
average emission rate during operation was 7.43 Ib/hr (Table 3-13).   The
average appears to represent the emissions at vent B.  The flows and
hydrocarbon concentrations were consistent over the three test runs.
Roof vent C is located above the dryers.  The average emission rate was
10.83 Ib/hr (Table 3-14).  The average appears to represent the emissions
at vent C.  The flows and hydrocarbon concentrations were consistent
over the three test runs.  Roof vent D is located above the end of the
dryer vent and conveyor belts.  The average emission rate was 6.03 Ib/hr
(Table 3-15).  The average appears to represent the emissions at vent D.
The flows and hydrocarbon concentrations were consistent over the three
test runs.  Comparing the emissions from all the vents, vent C consis-
tently produced more hydrocarbons.  The flow rates at each vent were
                                4
approximately the same (1.4 x 10  SCFM); therefore, the concentration of
hydrocarbons being emitted by the process around vent C are responsible
for the increased emission rate.  Vent C was located above the dryers
which were found to produce high concentration levels (Section 3.2.4).
The average emission rates from each vent appear representative of the
operation given the known process parameters.
     Emission flow rates, temperatures and moistures were measured at
each sampling locations.   Gas samples were collected following the
modified EPA Method 110 (Appendix G).  Analysis was performed by GC/FID
for the hydrocarbon concentration and by GC/TCD for the inert gas
concentrations.
3.2.4  Emulsion Crumb:   White Line Process:   Dryer Vents
     The rubber is carried from the presser to the dryers.   When dried,
the rubber is ready for pressing and packaging.  In the process building
there were two dryers with four vents each (Figure 5-10).  Some vents
are connected by ducting, resulting in two roof vents associated with
each dryer.  During the testing, only one dryer was in operation and
only one roof vent was operating.  The average emission rate for this
vent was 21.99 Ibs/hr (Table 3-16).  This average is based on the results
of two samples.  The results of each sample test were consistent but
cannot be considered representative of normal emissions from the dryers
because only two samples were taken, the production volume was low,  and
the other vents were inoperative.
                                 3-19

-------
                Table 3-13.   SUB-SUMMARY OF RESULTS AT  THE EMULSION
                       CRUMB  PROCESS  -  WHITE LINE ROOF VENT B
Ttit Run f
Date
Flue Gas Te*p (ฐF)
Flut Area (Ft2)
Flut Flow lUtc (SCFM)
Total Hydrocarbon
U1(i1on Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volune Percentages (X)
Hydrocarbon (as Cnpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
ERO-B-1 ERO-B-2 ERO-B-3 Average of Runs
6/2/80 6/3/80 6/3/80
93 105 98 99.0
15.155 15.155 15.155 15.155
1.55 x 104 1.56 x 104 1.56 x 104 1.56 x 104
7.48 6.71 8.11 7.43
As Benzene As Capd As Benzene As Capd As Benzene As Cซpd As Benzene As Cnpd '
—
—
—
0.3 0.4 N.A. N.A.
—
0.8 0.8 N.A. N.A.
—
10.8 10.8 N.A. N.A.
2.1 1.7 — — N.A. N.A.
—
—
39.9 34.9 33.9 29.6 32.4 28.4 35.4 31.0
39.9 34.9 36.0 31.3 44.3 40.4 35.4 31.0
0.00 0.00 0.00 0.00
21.60 21.40 21.00 21.33
78.10 78.00 78.00 78.03
0.00 0.00 0.00 0.00
99.70 99.40 99.00 99.37
N.A. - Trace amounts present In results but not averaged.
                                        3-20

-------
                    Table  3-14.   SUB-SUMMARY OF  RESULTS AT THE EMULSION
                             CRUMB  PROCESS  - WHITE  LINE ROOF VENT C
Ttst lun f
                                 ERO-C-1
                             EW-C-2
                       ERO-C-3
                                        Average of Runs
tote
Flw Cat TMP  (ฐF)
Flu* Art* (Ft2)
Flut Flow Rate (SCFH)
Total Hydrocarbon
Emission Rate  (U)/Hr)
         6/3/80

       •7
       15.155
        1.94 x 10*

        a. 65
  6/1/80

108
 15.155
  1.35 x 104

 10.81
  6/4/80

104
 15.155
  1.35 x 10*

 13.03
103
 15.155
  1.35 x 104

 10.83
Analytical Results (PPM)
As Benzene   As Capd  As Benzene   As Capd  As Benzene   As Cซpd  As Benzene    As Capd
Methane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethyl benzene
Xylene
Styren*
Total Hydrocarbon
~. ... ... ... ...
— ... ...
	 _ ซ. A 9
... ... ... ... g.ฃ
... 	 	 	
0.2
— — —
— ~. ... — 3.8
— 2.2 1.8
—
— — — ... ...
53.7 46.9 64.5 56.5 75.5
53.7 46.9 66.7 58.3 79.7
ป••
—
0.2
—
0.3
—
3.8
—
—
...
66.1
70.4
'.".'.
—
N.A.
—
N.A.
—
N.A.
N.A.
—
...
64.5
64.5
I"
—
N.A.
—
N.A.
—
N.A.
N.A.
—
...
56.5
56. 5
Volunt Percentages (X)
Hydrocarbon (as Cซpd)
Oxygen
Nitrogen
Water Vapor
0.00
21.50
78.00
0.00
0.01
20.90
78.00
0.00
0.01
21.00
77.80
0.00
0.01
21.13
77.93
0.00
Total  Percentage
       99.50
 98.91
                                                                           98.81
                                                                                                99.07
N.A.  • Trace amounts present in results but not averaged.
                                                  3-21

-------
                Table  3-15.  SUB-SUMMARY  OF RESULTS AT THE  EMULSION
                        CRUMB PROCESS -  WHITE LINE  ROOF VENT D
Ttst Run •
Date
F)ue CM Te*p (ฐF)
Flut Am (Ft2)
Flin Flew Rite (SCFH)
Total Hydrocarbon
Emission Rate (U>/Hr)
Analytical Results (PPM)
Hethane
Ethane
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Cnpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
ERO-D-1 ERO-D-2 ERO-D-3 Average of Runs
6/3/80 6/3/80 6/4/80
97 96 102 98
15.1S5 15.155 15.155 15.155
1.43 x 104 1.43 x 104 1.43 x 104 1.43 x 1C4
5.88 5.89 6.31 6.03
As Benzene As Capd As Benzene As Capd As Benzene As C*pd As Benzene As Dnpd
...
—
—
—
—
—
—
1.2 1.2 N.A. N.A.
—
... ... ... ... ... ... ...
—
33.0 29.7 33.9 29.6 35.2 30.8 34.0 30.1
33.0 29.7 33.9 29.6 36.4 32.0 34.0 30.1
0.00 0.00 0.00 0.00
21.40 21.00 21.00 21.13
78.10 78.00 78.00 78.03
0.00 0.00 0.00 0.00
99.50 99.00 99.00 99.17
N.A. - Trace amounts present In results but not averaged.
                                           3-22

-------
                Table  3-16.  SUB-SUMMARY OF RESULTS AT  THE EMULSION
                       CRUMB PROCESS -  WHITE LINE DRIER  VENTS
EDO-l EDO-2 EDO-3* Average of Runs
0ซU
Flu* 6ซl Teซp (ฐF)
Flut Art* (Ft2)
Flut Flew Rate (SCFN)
Total Hydrocarbon
Ulttlon Rat* (Lb/Hr)
Analytical RetulU (PPM)
Methane
Ethane
Propane
Butadiene
PtnUn*
HlMM
Cyclohexane
Benzene
Tolutne
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volune Percentage! (X)
Hydrocarbon (at Capd)
Oxygen
Nitrogen
Hater Vapor
Total Percentage
•74/60 C/4/BO
159 159 159
1.405 1.405 2.405
t.tt * 10S 5.88 x 10* 5.86 x 103
21.28 22.69 21.99
As Benzine At Oapd At Benzene At Capd At Benzene At Cซpd At Benzene At Caqid
_ ซ_ _ ^. _. ...
— — —
—.—-..- —
..._-_._. ซ..
— — — _
— — —
— — — —
3.5 3.5 2.5 2.5 3.0 3.0
1.4 0.8 1.7 0.9 1.6 0.9
3.9 2.7 3.9 2.7 3.9 2.7
— — — —
322.7 282.4 345.3 302.2 334.0 292.3
331.5 289.4 353.4 306.3 342.4 298.9
0.03 0.03 0.03
19.79 19.70 19.74
73.16 73.26 73.21
6.20 6.20 6.20
99.18 99.19 99.18
Vat not run due to incleaent Mather.
                                        3-23

-------
     Process data was collected by plant personnel.  Measurements were
made for flow rates, temperatures and moisture determination (Section 6.2.8).
A pseudo-stack had to be constructed at the vent to allow for sample and
data collection.  The gas samples were collected in accordance with
modified EPA Method 110 (Appendix G).  Analysis was performed on GC/FID
for hydrocarbon concentration and on GC/TCD for inert gas concentration.
3.2.5  Emulsion Crumb:  Black Master Roof Vents
     The black master line latex follows the same sequence of process
treatments as the white line latex; only the additives differ.   Therefore,
after leaving the storage tank, the latex flows through a blend tank, a
coagulation tank, a soap conversion tank, a water separator, a reslurry
tank, a presser and a dryer (Figure 5-13).   The black master line process
is conducted in a different building than the white line process
(Section 6.2.4).  There are four roof vents above the black master line
process (Figure 5-12).  Roof vent A is located above the coagulation
tank.  No samples or data were collected at A since the vent was not
operating during the testing period.   Roof vent B is centrally located
over the process at the pressers.   The average emission rate during
operation was .76 Ibs/hr (Table 3-17).   This figure is an average of
three tests.   The flow rate remained constant over the three tests, but
the concentrations of total hydrocarbon fluctuated from 67.6 ppm in the
first test to 18.4 ppm in the second test to 28.8 ppm in the third test.
These results may be attributable to ambient conditions and room flow
conditions at the time of testing.
     Roof vent C was located near the dryers (Figure 5-12).   The average
emission rate at vent C was 5.04 Ibs/hr (Table 3-18).   This emission
rate is considerably larger than the emissions from vent B.   The gas
flow at vent C was a factor of 10 larger than the flow at vent B, which
accounts for the variation.  The emission concentrations at vent C are
actually lower than the ones measured at vent B; for tests 1, 2 and 3,
the emission concentrations were 20.5 ppm,  24.7 ppm and 39.7 ppm,
respectively.
     Roof vent D was located above the dryers (Figure 5-12).   The average
emissions from the three tests was 6.44 Ibs/hr.   The flow rate  was
similar to the flow through vent C.   The concentrations measured was
                                 3-24

-------
                   Table 3-17.   SUB-SUfWARY  OF RESULTS  AT THE  EMULSION
                      CRUMB PROCESS -  BLACK MASTER LINE  ROOF VENT  B
Tart fern *
Data
Flu* CM Tea? (ฐF)
Flu* ArM (Ft2)
Flut Flow IUU (SCFN)
Total Hydrocarbon
blsslon Rate (Lb/Hr)
Analytical Results (PPH)
Methane
Ethant
Propane
Butadiene
Pontane
Nexane
Cyclohexane
Benzene
Toluene
Ettylbenzene
tyltne
Styrene
Total Hydrocarbon
VoluM Percentages (X)
Hydrocarbon (as Capd)
Oxygen
Nitrogen
Mater Vapor
Total Percentage
ERO-M-1 ERO-BM-2 ERO-BW-3 Average of Rum
6/13/BO 6/14/80 6/14/80
104 106 115 108
6.448 6.448 5.448 5.448
1.61 ป 1C3 1.61 x 10S 1.63 x 10S 1.62 x 103
1.22 .415 .649 .76
As Benzene As Capd As Benzene As Capd As Benzene As Capd As Benzene As Copd
_-___._.__.
_ — — _. — — —
— — - — — — — —
9.S 12. r — — — — K.A. N.A.
_~ — — — — — —
14.6 16.3 — — — — H.A. K.A.
21.9 21.9* 0.4 0.4 1.9 1.9 8.0 8.0
-— — — — — — — —
— — — — — —
3.8 2.6 2.6 1.8 3.8 2.6 3.4 2.3
— — - — — — —
16.7 14.6 18.6 16.3 27.8 24.3 21.0 18.4
66.3 67.6 21.5 18.4 33.5 28.8 32.4 28.7
0.01 0.00 0.00 0.00
20.4 20.4 20.4 20.4
77.3 76.6 76.6 76.8
0.0 0.0 0.0 0.0
97.7 97.0 97.0 97.2
*Value Is the sun of one or aore unknown compound! closely associated with the Identified compound.  Individual
 concentrations for each compound are  Included In the computer data sheets. Appendix A.

K.A. • Trace amounts' present In results but not averaged.
                                               3-25

-------
                  Table  3-18.   SUB-SUMMARY OF RESULTS  AT THE EMULSION
                     CRUMB PROCESS  - BLACK MASTER  LINE  ROOF  VENT  C
Ttlt Run f CRO-BMC-1
Date 6/13/80
Flut Cat Teejp (ฐF) 104
riuc Art* (Ft2) 10.896
Flue Flew Rate (SCFM) 1.34 x 104
Total Hydrocarbon
Emission Rate (Lb/Hr) 3.83
Analytical Result* (PPM) At Benzene Al Capd
Methane
Ethane
Propane — •"
Butadiene — . —
Pentane
Hexane —
Cyclohexane 0.8 0.8
Benzene — —
Toluene — —
Ethylbenzene 3.2 2.2
Xylene 0.5 0!3
Styrene 19.7 17.2
Total Hydrocarbon 24.2 20.5
Volume Percentages (X)
Hydrocarbon (as Cmpd) 0.00
Oxygen 20. 5
Nitrogen 77.3
Water Vapor 0.0
Total Percentage 97.80
Value Is the sum of one or more unknown compounds I
EfcO-WC-2 ERO-BMC-3 Average of Runs
fi/iJ/eO 6/14/80
lUi 102 103
10. ฃ96 10.896 10.896
).34 x 104 1 35 x 104 1.34 x 104
ซ.56 6.72 5.04
As Benzene As ฃmpd As Benzene As Capd As Benzene As Cmpd
—
... •

2.2 2.9 N.A. N.A.
...
3.9 4.4* N.A. N.A.
J.S 1.9 7.7 7.7" 3.5 3.5
...
2.9 2.3 N.A. N.A.
3.3 2.3 — — N.A. N.A.
N.A. N.A.
21.5 20.5 25.6 22.4 22.9 20.1
28.7 24.7 42.3 39.7 26.4 23.6
0.00 0.00 0.00
20.8 21.0 20.6
77.3 77.5 77.4
0.0 0.0 0.0
98.1 98.5 98.1
rlosely tssodated with the Identified compound. Individual
 concentrations for each compound are included in the computer data sheets.  Appendix A.
N.A. - Trace amounts present In results but not averaged.
                                             3-26

-------
slightly higher, which explains the higher emission rate at vent D.   For
the tests 1, 2 and 3, the emission concentrations were 30.7 ppm, 25.5
ppm and 30.5 ppm, respectively (Table 3*19).
     Emission flow rates, temperatures and moisture determinations were
measured at each vent.  Gas samples were collected in accordance with
the modified EPA Method 110 (Appendix G).  Analysis was performed by
GC/FIO for the hydrocarbon concentrations and by GC/TCO for the inert
gas concentrations.
3.2.6  Emulsion Crumb:  Black.Master Line:  Dryer Vents
     The dryers are the final treatment step in the production process.
The crumb enters the dryers from the presser and are pressed and packaged
at the end.  In the black master line process there were two dryers,
each with four roof vents (Figure 5-12).  During the testing, both
dryers were operating.  Samples were collected from two vents off each
dryer.  Vents A and B were sampled from one dryer.  Vents C and 0 were
sampled from the other dryer.  The total emissions (average) from vents
A and B were 13.96 Ibs/hr (Table 3-20 and 3-21).   The total emissions
(average) from vents C and "0 was 16.49 Ibs/hr (Table 3-22 and Table 3-23).
Three samples were collected from each vent.  The average emissions were
calculated for each vent from three samples with the exception of vent A.
     At vent D, both the flow rates and emission concentrations were
consistent with all three tests (Table 3-23).  The average emission rate
of 10.66 Ibs/hr is considered representative of the conditions of the
process during the sampling period.
     In general, the results cannot be considered as representative
because the process was not operating at normal operating conditions and
only two vents per dryer were operating.  However, plant personnel
Indicate that although the hourly rate of emissions might be greater at
full load capacity, emissions per pound of rubber produced would likely
decrease because of reduced dryer efficiency at the higher rates.
                                 3-27

-------
                   "Table 3-19.   SUB-SUMMARY OF RESULTS AT THE EMULSION
                       CRUMB  PROCESS  - BLACK MASTER LINE ROOF  VENT D
Ttit Run f
Date
Flue Gas TMP (ฐF)
Flue Arta (Ft2)
Flue Flow Rate (SCFH)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Nexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Cซpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
ERO-BHD-1 ERO-BMD-2 ERO-BHD-3 Average of Runs
6/13/80 6/14/80 6/14/80
104 105 115 108
10.896 10.896 10.896 10.896
1.73 x 104 1.73 x 10* 1.73 x 10* 1.73 x 10*
6.42 5.67 7.24 6.44
As Benzene As Capd As Benzene As Cnpd As Benzene As Cซpd As Benzene As Cmpd
... ... ... ... ... ... ... —
—
—
2.6 3.4 1.3 1.7 — — N.A. N.A.
... ... ... ... ... - — ... —
6.2 7.0a 1.6 1.8* — — N.A. N.A.
5.5 5.5 2.7 2.7 0.6 0.6 3.0 3.0
—
2.4 2.0 N.A. N.A.
2.7 1.9 2.3 1.6 — — 1.7 1.2
—
14.7 12.9 20.3 17.7 31.7 27.9 22.2 19.5
31.7 30.7 28.2 25.5 34.7 30.5 25.2 22.5
0.00 0.00 0.00 0.00
20.7 21.1 20.8 20.9
77.7 78.4 77.2 77.8
0.0 0.0 0.0 0.0
98.4 99.5 98.0 98.6
'value Is the sum of one or mart unknown compounds closely associated with the Identified compound.  Individual
 concentrations for each compound are Included 1n the computer data sheets, Appendix A.

N.A. - Trace amounts'present in results but not averaged.
                                                 3-28

-------
                  Table 3-20.   SUB-SUMMARY OF RESULTS AT THE  EMULSION
                    CRUMB PROCESS - BLACK MASTER  LINE DRYER VENT A
Ttit Bun f
Date
Flue fiat TMP (ฐF)
riut Area (Ft2)
Flut Flow Rate (SCFH)
Total Hydrocarbon
Emission lUtc (U/Hr)
Analytical Result* (PPM)
Nethane
Ethan*
Propane
Butadiene
Pentane
Hexane
Cyclohexane
Benzene
Toluene
Ethyl benzene
Xylene
Styrtne
Total Hydrocarbon
Voluae Percentages (%}
Hydrocarbon (as Capd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
EDO-BNA-1 EDO-BMA-2* CDO-BNA-3 Average of tuns
6/13/80 6/14/80 f/14/M
152 153 153 153
7.K6 7.126 7.826 7.826
C.78 * 10S C.82 ซ 103 C.76 ซ 10* C.79 x 10*
10.67 4.87 11.45 11.06
As Benzene As Capd As Benzene As Ctapd As Benzene As Cspd As Benzene As Capd
_—_._ — .„.„___„
— — — — —. ^.
~. — — — — —
1** 1ป4 ••• ซ^ป ••• •*• M. A. N. A.
— — ~. — — —
1.7 1.9 — •• *•• — • ••• H. A. N. A.
3.4 3.4 — * — • •— • -•*• H.A. N.A.
— ... ... — _ ...
— — 1.0 0.8 1.6 1.3 N.A. N.A.
2.3 1.5 — — 1.7 1.2 2.0 1.4
ซ M A • ••• —---— — ^_ — -_ -— MA tl A
A.^ *v ••• •ป — — _v~ n.n. n. n.
131.2 115.6 62.4 55.0 146.2 128.8 138.7 122.2
141.1 124.7 62.4 55.8 149.5 131.3 140.7 123.6
0.01 0.01 0.01
18.76 18.67 18.76
70.48 70.94 70.11
8.47 8.47 8.47
97.72 98.09 97.35 97.72
*8ปg had a leak, analyzed but results ncre not Included In the average.
N.A. - Trace eaounts present In results but not averaged.
                                             3-29

-------
                 Table 3-21.   SUB-SUMMARY OF  RESULTS AT THE EMULSION
                   CRUMB  PROCESS  -  BLACK MASTER LINE DRYER VENT  B
Ttft Run *
Oat*
Flu* Gat Teap (ฐF)
Flu* Arta (Ft2)
Flu* Flow Rat* (SCFH)
Total Hydrocarbon
Emission Rate (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pentane
Nexane
Cyclohexane
Benzene
Toluene
Ethylbenzene
Xylene
Styrene
Total Hydrocarbon
Volume Percentages (X)
Hydrocarbon (as Cซpd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
EOO-8MB-1 EDO-BHB-2 EOO-BMB-3 Average of Runs
6/13/80 6/13/80 6/13/80
134 124 130 129
5.444 $.444 S.444 5.444
4.05 x 103 4.05 x 103 4.1 x 103 4.07 x 103
5.09 4.42 5.37 4:96
At Benzene At Capd -As Benzene As Capd At Benzene At Cซpd At Benzene At Cupd
— — — —
— — —
...
... ... ... ... — ... —
—
—
0.9 0.9 — — 0.6 0.6 N.A. N.A.
—
2.5 2.1 1.5 1.2 4.0 3.3 2.7 2.2
1.1 0.7 	 — 1.3 0.9 N.A. N.A.
i
103.8 90.8 90.6 79.8 107.0 94.3 100.5 88.3
108.3 94.5 92.1 81.0 112.9 99.1 103.2 90.5
0.01 0.01 0.01 0.01
18.97 18.97 19.15 19.03
70.90 69.98 71.27 70.72
7.92 7.92 7.92 7.92
97.80 96.88 98.35 97.68
N.A. - Trace amounts present in results but not averaged.
                                           3-30

-------
                   Table 3-22.   SUB-SUMMARY  OF  RESULTS  AT  THE  EMULSION
                      CRUMB PROCESS -  BLACK MASTER  LINE  ROOF VENT  C
Tซlt Run •
Date
Fit* fias TMO (ฐF>
Flut Area (Ft2)
Flut Flew Rate (SCFN)
Total Hydrocarbon
Emission Ratt (Lb/Hr)
Analytical Result* (PPM)
Ntthanc
RlMM
Propane
Butadltnt
PenUne
Hexane
Cyclohexane
Btnxtnt
Tolutn*
Cthylbenzene
Xylene
Styrtnc
Total Hydrocarbon
VoluM Percentage* (X)
Hydrcarbon (as C>pd)
Oxygen
Nitrogen
ttater Vapor
Total Percentage
EDO-BMC-1 EDO-BMC-2 EOO-BMC-3 Average of Runt
i/13/BO C/U/BO f/13/80
119 N 103 107
3.276 3.276 J.276 1.276
S.M * 10S 6.54 x 10S 5.S6 x 10S S.54 x 103
7.76 4.24 5.48 5.83
As Benzene As Cซpd As Benzene As Cepd As Benzene As Cepd As Benzene As Copd
— ~. ~_ ~. — _.
— ~. — . — ~. ซ.. ^.
— ~. ~. ~. — —
— — 0.3 0.5 1.2 1.9 H.A. H.A.
___.___-
0.7 O.B 0.6 0.7 3.5 3.9* 1.6 1.6
1.9 1.9 1.0 1.0 3.4 3.4 2.1 2.1
— — -——- -~ — —
— — 3.8 3.1 5.7 4.7 N.A. N.A.
2.8 1.9 — — 1.1 0.7 N.A. N.A.
— — — —.•—-. — -
107.6 94.2 57.5 50.7 67.5 59.5 77.5 68.1
113.0 98.8 63.4 56.0 82.4 74.1 81.2 72.0
0.01 0.01 0.01 0.01
19.88 19.88 20.08 19.95
72.00 74.42 74.90 73.77
3.48 3.48 3.48 3.48
95.37 97.79 98.47 97.21
*Value is the sun of one or mort unknown conpoundt closely associated with the Identified compound.  Individual
 concentrations for each compound are included in the computer data sheets. Appendix A.
N.A. - Trace anounts. present in results but not averaged.
                                                3-31

-------
               "Table 3-23.   SUB-SUMMARY OF  RESULTS AT THE EMULSION
                   CRUMB  PROCESS -  BLACK MASTER LINE ROOF  VENT D
Ttst Run •
Oat*
Flue CM Imp (ฐF)
Flut Area (Ft2)
Flue Flow Rat* (SCFN)
Totซl Hydrocarbon
Emission Rat* (Lb/Hr)
Analytical Results (PPM)
Methane
Ethane
Propane
Butadiene
Pcntane
Hexane
Cyclohexane
Benzene
Toluene
Ethyl benzene
Xylene
Styrene
Total Hydrocarbon
. Volume Percentages (X)
Hydrocarbon (as Copd)
Oxygen
Nitrogen
Water Vapor
Total Percentage
EDO-BMD-1 EDO-BMD-2 EDO-BMD-3 Average of Runs
6/13/80 6/13/80 6/14/80
142 139 137 139
5.526 5.526 5.526 5.526
5.92 x 103 5.87 x 103 5.92 x 103 5.90 x 103
10.07 11.44 10.47 10.66
As Benzene As Qapd As Benzene As Capd As Benzene As Capd As Benzene As Cnpd
—
—
—
0.5 0.8 0.7 1.2 N.A. N.A.
—
0.8 0.9 1.4 1.5 N.A. N.A.
0.8 0.8 1.8 1.8 2.8 2.8 1.8 1.8
—
5.7 4.6 5.1 4.1 N.A. N.A.
2.5 1.7 2.3 1.6 1.8 1.3 2.2 l.S
1.5 1.0 — — — — N.A. N.A.
143.7 • 125.8 157.1 138.4 141.0 124.2 147.3 129.4
148.5 129.3 168.2 148.1 152.86 135.1 151.3 132.7
0.01 0.01 0.01 0.01
18.79 18.89 18.89 16.86
70.59 69.58 69.77 69.98
8.32 8.32 8.32 8.32
97.71 96.80 96.99 97.17
N.A.  - Trace amounts present In results but not averaged.
                                          3-32

-------
                        4.0  PROCESS DESCRIPTION

     Phillips Petroleum's copolymer plant contains two production lines
used in manufacturing styrene-butadiene rubber:   the Emulsion Crumb
Process and the Solution Crumb Process.  A number of different products
are produced by each process.  This section will discuss the Emulsion
Crumb Process with emphasis on emission points and control  devices.
Because of the alleged proprietary nature of the Solution Crumb Process,
this process is not discussed in this section of the report.
     The basic difference between the two processes are the solvents
used during the manufacturing and finishing of the product.   The solution
process uses an organic soWent while the emulsion process  is water-based.
An attempt was made during the sampling trip to characterize most of
these emissions.  Not all points of emissions were tested due to time
constraints; however, a sample was taken at each major point that was
representative of the plant process.   The following descriptions are
based on normal plant operation and,  in some cases, differ  from what  was
observed during testing.
4.1  THE EMULSION CRUMB PROCESS (Figure 2-2)
     The emulsion process contains four major lines of SBR  products.
Three of these lines are for the production of "Black Master" SBR and
one is for "White" SBR.  Production of all lines precedes basically the
same up until the blend tanks in the processing section (see Figure 1-1).
After the blend tanks in the black line, oil and carbon black are added
to the SBR latex to give increased strength and durability.   After this
addition the lines are again processed alike.
     Part A (Manufacturing) and Part B (Processing) are located in
separate buildings at the plant and are separated in this report for
clarity.  The latex is formed in the reactors from raw compounds

-------
(styrene and butadiene).  Upon completion of the reaction, the latex is
routed through the blowdown tanks to the flash tank, which is under
vacuum.  In the flash tank most of the residual butadiene is flashed off
due to its high volatility.  The butadiene is processed through the
pressure recovery tank, the vacuum recovery tank and the brine chiller
In order to condense and recover the VOC's from the latex.  The uncondensed
gas emissions are passed through a kerosene absorber to further reduce
the emissions, and after the liquid kerosene is removed, vented to the
atmosphere.
     The latex leaves the flash tanks and enters the stripper where
steam stripping removes the residual styrene which is then recovered at
the chiller.  There are three lines in the manufacturing of latex.  The
Phillip's designations are A, B, and E.   Line A and B are older (circa
1941) and smaller than line E (circa 1953).   The stripper on line E is
approximately twice the size of the strippers on lines A and B.   After
stripping, the latex is stored for three to four days until the processing
section has need for it.
     The processing begins Vith the latex being blended to product
specifications.  The latex is added to a water slurry to coagulate the
latex and is then solidified in the soap conversion tank.   The water/latex
slurry is separated and the latex crumb washed in the reslurry tank.
The wash water is removed in the presser.  The crumb is transferred to
the drier where it is heated to drive off the remaining moisture.   The
dried crumb is then pressed and packaged for shipment.
     During the testing period there was a low demand for the SBR product.
Both "White" and "Black Master" lines were operating at reduced levels.
There was only one white line in operation,  and it was operating at 60%
of its normal  maximum capability.   The Black line was at normal  operating
levels, however, only one line was in operation at the time of the test.
4.2  EMISSION/PRODUCTION RATE
     The emission/production rate at the Borger Facility was calculated
from emission results of the test locations  compared to the production
rate through the process locations during testing.   The emission rate is
the pounds per hour mass rate from the test results.   The production
                                 4-2

-------
rate through the process Is the flow data provided by the Phillips'
Petroleum personnel (Appendix D).   Table 4-1 presents the data utilized
for the emission/production rate calculation.
                                 4-3

-------
TABLE 4-1.  EMISSION/PRODUCTION RATE

Test #


ERO-A-1
A-2
A-3

ERO-B-1
B-2
B-3

ERO-C-1
C-2
C-3

ERO-D-1
D-2
D-3

EDO-1
-2

ERO-BMB-1
-2
-3
Day


6/2
6/3
6/4

6/3
6/3
6/4

6/3
6/3
6/4

6/4
6/4
6/4

6/4
6/4

6/13
6/13
6/13
Emission Rate
Ibs/hr

5.94
3.67
12.47

7.48
6.71
8.11

, 8.65
10.81
13.03

5.88
5.89
6.31

21.28
22.69

1.22
0.41
0.65
Production Rate
Ibs/hr

5717
5973
5100
(Average)
5973
5973
5100
(Average)
5973
5973
5100
(Average)
5100
5100
5100
(Average)
5100
5100
(Average)
12420
12420
12420
Emission per
Production
(Ibs/lbs)
1.04 x 10"3
6.14 x 10"4
2.44 x 10"3
1.36 x 10"3
1.25 x 10"3
1.12 x 10"3
1.59 x 10"3
1.32 x 10"3
1.45 x 10"3
1.81 x 10"3
2.55 x 10"3
1.94 x 10"3
1.15 x 10~3
1.15 x 10"3
1.24 x 10"3
1.18 x 10"3
4.17 x 10"3
4.45 x 10"3
4.31 x 10"3
9.82 x 10"5
3^30 x 10"5
5.23 x 10"5
                                            6.12 x 10
                                                     -5
               4-4

-------
Table 4-1.  Continued
Test f


ERO-BMC-1
-2
-3

ERO-BMD-1
-2
-3

EDO-BMA-1
-2
-3

EDO-BMB-1
-2
-3

EOO-BMC-1
-2
-3
Day


6/13
6/13
6/13

6/13
6/13
6/14

6/13
6/14
6/14

6/13
6/14
6/14

6/13
6/14
6/14
Emission Rate
Ibs/hr

3.83
4.56
6.72

6.42
5.67
7.24

10.67
4.87
11.45

5.09
4.42
5.37

7.76
4.24
5.48
Production Rate
Ibs/hr

12420
12420
12420
(Average)
12420
12420
12375
(Average)
12420
12375
12375
(Average)
12420
12375
12375
(Average)
12420
12375
12375
Emission per
Production
(Ibs/lbs)
3.08 x 10"4
3.67 x 10"4
5.41 x 10~4
4.05 x 10"4
5.17 x 10~4
4.56 x 10"4
5.85 x 10"4
5.19 x 10"4
8.59 x 10"3
3.93 x 10"3
9.25 x 10"3
7.26 x 10"3
4.10 x 10"3
3.57 x 10"3
4.34 x 10~3
4.00 x 10"3
6.25 x 10"3
3.43 x 10"3
4.4 x 10"3
                        (Average)  2.77  x  10"3
     4-5

-------
Table 4-1.  Continued
Test #


EDO-BMD-1
-2
-3

SCO-1
-2
-3

SHO-1
-2a
-3

SYO-1
-2
-3

SDO-A-1
-2
-3
Day


6/13
6/14
6/14

6/10
6/10
6/10

6/10
6/10
6/10

6/5
6/5
6/10

6/5
6/5
6/10
Emission Rate
Ibs/hr

10.07
11.44
10.47

13.91
15.17
25.62

2.73
' 0.96
1.98

12.12
9.59
7.12

40.23
22.65
69.06
Production Rate
Ibs/hr

12420
12375
12375
(Average)
6775
6775
6775
(Average)
6775
6775
6775
(Average)
6692
6692
6775
(Average)
6692
6692
6775
Emission per
Production
(Ibs/lbs)
8.11 x 10"3
9.24 x 10"3
8.46 x 10"3
8.60 x 10"3
2.05 x 10"3
2.24 x 10~3
3.78 x 10~3
2.69 x 10"3
4.03 x 10~4
1.42 x 10"4
2.92 x 10"4
3.47 x 10"4
1.81 x 10"3
1.43 x 10"3
1.05 x 10~3
1.43 x 10"3
6.01 x 10"3
3.38 x 10"3
1.02 x 10"2
                        (Average) 6.53 x 10
                                          -3
    4-6

-------
                            Table 4-1.   Concluded
 Test *
Day      Emission Rate
            Ibs/hr
Production Rate
    Ibs/hr
Emission per
 Production
 (Ibs/lbs)
SDO-B-1
-2
-3

SDO-C-1
-2
-3

SDO-D-1
-2
-3

SDO-E-1
-2
-3

6/5
6/5
6/10

6/5
6/5
6/10

6/5
6/5
6/10

6/5
6/5
6/10

14.96
4.99
29.20

1.34
0.92
12.60

1.16
i
' 0.29
4.90

0.23
0.83
3.04

6692
6692
6775
(Average)
6692
6692
6775
(Average)
6692
6692
6775
(Average)
6692
6692
6775
(Average)
2.24 x 10"3
7.46 x 10"4
4.31 x 10"3
2.43 x 10"3
2.00 x 10"4
1.37 x 10"4
1.86 x 10~3
7.32 x 10"4
1.73 x 10~4
4.33 x 10"5
7.23 x 10"4
3.13 x 10"4
3.44 x 10"5
1.24 x 10"4
4.49 x 10"4
2.02 x 10"4

Unrepresentative test run and not included  in average.
                                 4-7

-------
                          5.0  SAMPLE LOCATIONS

     This section of the report will describe the location of sample
points in the processes at the Phillips Petroleum Borger facility.   The
locations discussed will be divided into the two main processes and
further subdivided according to their position along the process path.
The labeling scheme for the sample location is presented in the glossary,
and the plant layout with sample locations is presented in Figures  2-1
and 2-2.
5.1  SOLUTION CRUMB PROCESS
     This process has sample locations at the atmospheric vents of  the
process tanks, the liquid sampling valves of the latex stream, and  the
exhaust vents from the production processing system.   The sample locations
will be discussed in separate sections.
5.1.1  Flash Tank Ammonia Condenser (Figure 5-1)
     The gaseous flow from the flash tank was processed by an accumulator
and then by an ammonia condenser.  Sample locations were stationed  at
the inlet (SFI) and the outlet (SFO) of the ammonia condenser.  The
modified SFI position was sampled from a nitrogen feed stream which
entered the flash tank accumulator.  Sampling was accomplished by shutting
off the nitrogen feed to allow outlet venting of the flash tank accumulator.
This sample was representative of the inlet to the ammonia condenser.
The sample was obtained by the modified integrated bag sampling according
to the procedure outlined in the EPA Federal Register Method 110.   The
sample point was positioned after the valve to control the bleed-off
from the accumulator.  The bleed-off piping was reduced to a V connector
for direct attachment to the sampling train.  The sample flow to the
train was controlled by the ball valve in the bleed-off line and a  pinch
clamp placed on the sample train probe.  The sample flow and temperature

-------
     To 'Sampling
     Apparatus
     At Sample-^*
     Point SFO
        1"  to Pipe
        Feed
Inlet
From
Flash
Tank
                           Pressure Activated Valve
                                5" H0
                                                  Annonla Out
                                 Gas Phase
                              Inlet From Flash Tank
                                Liquid Phase
                                                           Bleed Valve
                                                          Level Gauge
                                                         SIDE VIEW
                                  To Process
    Figure  5-1.
Flash tank  accumulator with ammonia  condenser:

 sample points SFI and SFO

               5-?

-------
were not monitored from the modified SFI sample position because the
conditions at this sample point was considered unrepresentative of the
Inlet to the ammonia condenser.  For this reason, the inlet flow and
temperature was assumed to be the same as the outlet flow from the
condenser.
     The SFO position had a direct outlet to the atmosphere.   The
condensation in this stream created problems in obtaining a gaseous
sample at this point.  Therefore, a modified sampling apparatus (see
Figure 5-2) was designed to accommodate the liquid in the stream and the
annubar for flow measurements.  The sampling apparatus captured the
liquid condensate in a catch tank and the gaseous flow was discharged
through a 2-inch pipe into the atmosphere.   The liquid condensate amount
in the catch tank was measured after each sample run and a composite of
the three sample runs was obtained for analysis.  The gaseous flow from
the modified sampling apparatus was monitored by placing a FTL-75 Dieterich
Annubar Flowmeter in-line with the 2-inch pipe emitting to the atmosphere.
The annubar was placed 13 inches from any Interference upstream and
10 inches from the gas exit port to the atmosphere.   The sample point
for the modified Integrated bag sample was 5 inches  inside of the pipe
at the gas exit port.  This sample point was obtained by securing the
probe from the sample system 5 inches inside of the  pipe.   The temperature
was also monitored at this point by securing the thermocouple wire
directly to the sample probe.  The evacuated can sampling system was
used at the SFO sample point to obtain the gaseous sample.  The moisture
content of this stream through the condenser was assumed to be zero
because the liquid condensate from the stream was analyzed as 100%
hydrocarbons.
     The sampling procedure for analyzing the gas stream through the
ammonia condenser was preceded by monitoring the process for reactor
dumps.   The outlet pressure valve would open when the reactor dump
began.   The inlet and outlet sample system were initiated as the valve
opened.  The flows to the sample bags were maintained at a constant rate
with pinch clamps for adjusting on the sample line.   The sample systems
remained on-line until the gas flow dropped off and remained down.  The
initial high flow would last approximately five minutes and then drop
                                 5-3

-------
               1" ID Pipe
                     Fran *non1* Condcnicr Mปvt Flilh Tank
                                              To  Sample System and
                                              Temperature  Readout
                                                              Gas Flow
                                   Air Tight Catch Tank
Figure  5-2.  Sampling apparatus at SFO

                       5-4

-------
off, but would fluctuate for another minute.   The sample remained on-line
until the fluctuation stopped.  The gas sample would then be recovered
and transported to the on-site lab for analysis and the liquid condensate
was measured for volume.
5.1.2  Blend Tank Ammonia Condenser (Figure 5-3)
     At the modified inlet sample location (SBI), gaseous emissions
exited from a two-way pressure cap on the top of the blend tank.   A
sample probe was inserted directly into this port during a positive flow
from the blend tank.  This flow represented the concentrations from the
blend tank headspace that would flow through the control device under
regular operating conditions.
     The gaseous sample was obtained with the evacuated can system for
sampling according to the Modified EPA Method 110.  The sample point was
10 inches below the pressure cap.   The gas sample and temperature were
obtained at this point by inserting the sample probe inside the tank
with the thermocouple wire secured to the probe.  The flow during the
pressure cap release was inconsistent and a flow measurement was
unobtainable.             '
     The procedure for analyzing the modified ammonia condenser inlet
was to observe the process for the filling of a blend tank.   The pressure
cap on this tank would open when the headspace vapor reached 5 inches of
water.  The sample system probe would be inserted and the sample extracted
from inside the tank.  The sample system was operated until  the cap
resealed.  The sample bag was recovered and transported to the mobile
lab for analysis.
5.1.3  Steam Stripper
     The latex flow from the blend tanks through the steam strippers was
tested.  Steam stripper inlet (SSI) samples were taken directly from a
plant sample by-pass valve located on the bottom of the blend tank.  The
outlet sample location (SSO) was the crumb stream exiting the stripper
enroute to the crumb tank.  Samples were obtained from a plant sampling
valve.
                                 5-5

-------
                                     Pressure
                                      Gauge
                   From Other Blend Tanks
Probe to Sample System
 from Sample Point SBI
 Sample Point SSI _
                            BLEND
TANK
                                                                  Pressure Activated
                                                                  Valve
                                                                        Ammonia
                                                                        Condenser
                                                                   From Flash Tank
                                                                    To Cement Steam
                                                                    Stripper
                Figure 5-3.  Blend tank wnmonia  condenser
                          sample points SBI  and  SSI
                                            5-6

-------
     The latex sample was in a cement phase at the Inlet stage of the
stripper.  The sample point was taken from the blend tank by-pass valve
that was flowing Into the steam stripper.   The sample was obtained from
the by-pass valve according to the sample  procedure of the EPA draft
method Determination of Residual Hexane in Solution Crumb Rubber by Gas
Chromatographic Analysis (Appendix F).  The method for measuring the
volume of sample was developed because the high volatility of the cement
phase made a volumetric measurement unobtainable.   Therefore, the volume
of sample was determined by gravimetric analyses.   The adapted method of
utilizing the density and weight of the sample for volume determinations
is described in Section 6.  The temperature was taken directly from the
cement sample bottle with a metal thermometer.
     The latex sample was in a crumb form  combined with water at the
outlet stage of the stripper.  The sample  point was taken directly from
a sample valve exiting the stripper to the crumb tank.   The method of
sampling the crumb was according to the sampling procedures of the EPA
Draft Method: Determination of Residual Hexanes in Solution Crumb Rubber
(Appendix F).  The temperature of the solution was taken by direct
measurement with a metal thermometer from  the water extracted from the
crumb.
5.1.4  Crumb Tank Vent (Figure 5-4)
     The vent directly above the headspace of the crumb tank was maintained
as a sample location (SCO).  Crumb from the stripper remained at high
temperatures and emitted gaseous effluents, therefore, the space within
the tank was constantly monitored by plant personnel for explosive
concentrations.  An ambient air stream was flushed through the tank to
maintain a dilution of the high concentrations.  Samples were collected
through a Teflon probe inserted 3 feet into the 10-foot stack (see
Figure 5-4).  The three feet inside the stack was maintained as the
sample point.  The evacuated can system was utilized in obtaining a
modified Method 110 gas 'sample.  The temperature of the gas stream was
monitored during testing by taping a thermocouple wire to the end of the
probe inserted into the stack.  The flow from the stack was measured
with an anemometer.  The flow measurements were taken at the center of
stack exit and averaged.  A standard midget impinger train was used for
                                 5-7

-------
 Plant
Explosion
 Meter
                  Gas Flow
                       ti
                       1
                           .33"
                      \
Probe to Sample Train (Sample point  SCO

                    3 Ft.  Inside  Stack)
       Agitator
                        Open for
                        Excess Ambient
                        A1r Flow
                                                                Manhole
      Figure  5-4.  Crumb slurry  tank outlet:  sample point  SCO
                                          5-8

-------
moisture determination and sampled according to EPA Method 4 procedures
from the sample point Inside the stack.
5.1.5  Dewatering Screen Hood and Cyclone Vent (Figure 5-5)
     The dewatering screen process separates water from the crumb.   The
crumb product is conveyed into an extruder and pelletizer.  While the
water from this procedure was recycled to the stripper outlet, gas
emitted from the separation process was vented to the atmosphere.  The
gas effluent was captured above the screens by a hood, and fan-forced to
a roof vent.  The sample location (SHO) was the roof vent above the
hood.  A pseudo-stack (see Figure 5-5) was designed for the sample point
and adapted to the roof vent to allow unrestricted flow measurements and
collection of integrated bag samples.
     The crumb product conveyed from the dewatering screen and extruder
was processed through a pelletizer.  Pellet products then were blown
through an air stream cyclone.  The cyclone allowed the pellets to fall
out of the stream and into the dryer system, while any gas emitted from
this process remained in the cyclone flow.  The cyclone flow was vented
to the roof and a sample location (SYO) was maintained for the collection
of gas emissions.  Samples were taken directly from the vent for integrated
gas samples, but the flow measurements required a pseudo-stack for
proper flow determinations.
     Both sample locations required the evacuated can system for the
extraction of the modified EPA Method 110 integrated bag sample.  The
sample point for SHO was from the sample port on the pseudo-stack.   The
probe was inserted into the sample port at the middle of the pseudo-stack.
The temperature of the gas stream was also monitored at this location by
securing the thermocouple to the sample probe.  A midget impinger train
was utilized at the sample point for moisture determination according to
EPA Method 4.  The flow measurements required the pseudo-stack to give
the standard conditions for using the EPA Method 1 procedures of gas
velocity determinations with a S-type pitot tube (see Figure 5-6).
     The sample point for SYO was positioned 3 feet inside the exit of
the stack.  The cyclonic flow caused problems with maintaining the same
position.  But this problem was minimized by securing a weight to the
sample probe.  The temperature of the gas stream was monitored at the
                                 5-9

-------
            Pseudostack Adapted To Vent
 I
t->
o
                                   Sample Point SYO       San!p1e Point SD0.A

                                              Drier Vent
Crumb and
Hater
Solution  ,
                                            BOO
                                                       DOOQ
                                                                  9000
I  oo I    I oo I    I oo I
           DewateHng
           Screen  .
                                            Blower
Drier Vent
A
B
C
D
E
Temperature (ฐF)
144
186
196
193
189
                                             Conveyor
                                                               Drier
                                                                        Roof
                                                                                                   Finished
                                                                                                   Product
                                  Figure 5-5.   Solution process  -  SHO,  SYO,  SDO's

-------
(71

. Traverse Point
' Number
A*B 1
2
3
4
5
6
7
8
9
10



Traverse Point
Location From
Outside of Port
(Inches)
.351
1.107
1.917
3.051
4.617
-- 8.883
10.449
11.529
12.393
13.149
(Dewatering
Screen
Vent)
                                                                Fan
                                                                 \
   ~   (Side View)

  Pseudo Stack
Probe to Sample System
from Sample Point SHO


   (Traverse  Points)
                                                                  90'
                                                                                      {-  	I 13.5
                                       Figure 5-6.  Dewatering screen hood vent:   sample point SHO

-------
sample point by securing the thermocouple to the sample probe.  This
sample point was utilized for the moisture determination according to
EPA Method 4.
     The flow measurements required the pseudo-stack adaptation to
provide the standard conditions for using the EPA Method 1 procedures of
gas velocity determinations with a standard pi tot tube (see Figure 5-7)
and angle procedures for the measurement of cyclonic flow.
5.1.6  Dryers
     Gas emissions from the dryer system were vented to the roof and
directed into the atmosphere.  The dryer had five roof vents (SDO) along
the system.  The overall diagram of the vent positions are presented in
Figure 5-5.  The SDO sample locations required the evacuated can system
for the extraction of the modified EPA Method 110 integrated bag sample.
The sample point was 6 inches inside the sample ports (see Figure 5-8)
at the base of the dryer vents.   This point was maintained by inserting
the sample probes for the gas sample and moisture train inside one of
the sample ports.   The moisture determination procedure was according to
EPA Method 4.  The temperature was monitored by securing the thermocouple
to the gas sample probe.  The flow procedures, according to EPA Method 1,
required the addition to the dryer vents of sample ports (Figure 5-8).
The sample ports were installed and the gas velocity measured with a
S-type pitot tube.
5.2  EMULSION CRUMB PROCESS
     This process had sample locations at the outlet vent above the
kerosene knockout tank (combined emission stream from the flash tank and
styrene stripper system), the latex flow through the steam stripper
system, the fugitive emission roof vents, and the dryer vents.   The
black master and white line involved in the process had duplicate sample
locations.   The sample locations of the Emulsion Crumb Process are
presented in the process diagram (Figure 2-2) and will be discussed
separately in this section.
5.2.1  Kerosene Absorber (Figure 5-9)
     Gaseous outlet streams venting from the flash tank reaction and the
styrene strippers were processed through recycling devices, and the
                                 5-12

-------
                                                  (Side View)
Traverse Point
Nu*er
A* 1
2
3
4
5
6
7
8
9
10
11
12


Traverse Point
Location Fron
Outside of Port
(Inches)
.4
1.3
2.2
. 3.3
4.7
6.7
12.1
14.1
1S.4
16.5
17.5
18.4
(Cyclone Outlet
Vent)
                                                 Pseudo-Stack
                                                       Probe to Sample Systea
                                                       fro* Sample Point STO
                                          18.75"
                                                        ,r-  (Traverse Points)_
Figure 5-7.  Cyclone vent:   sample point SYO

-------
(Rear View)
Traverse
Point Number
A-E 1
2
3
4
Traverse Point
Centroid Lo-
cation (Inches)
1.7
5.1
8.5
11.9
(Solution Dryer
Vents)
                                             (Vents)
                                             1
                                             B
        1    1
                                                                                 (Side View)
                                                               13 1/2"
                                             40  1/2"


•

5
(




11 1C
Trave



It tL


•





r t r -
10" 5"
rse Points)
                                                                                   Probe  to  Sample
                                                                                   System from Sample
                                                                                   Point  5DO-X
                Figure  5-8.   Solution crumb process dryer vents:  sample points
                             SDO-A, SDO-B, SDO-C, SDO-D, and SDO-E.

-------
tn
•-•
in
                 Level
                 Gauge
              -20'
                   Recycled  to
                 Kerosene Absorb
                                                   r
                                                      '35'
                                                           Vent
                                               Sample Point EKO
Kerosene Knockout Tank
                                                                    Oriffce
                                                                                From Kerosene
                                                                                  Absorber
                                            D/P Gauge
                                                                                              Sample
                                                                                            Apparatus
                     Ground
                          Figure  5-9.   Emulsion  process  kerosene  tank:  sample point  - EKO.

-------
remaining gas stream was charged through a kerosene absorber.  The
sample location for this gas stream was to be at the exit of (EKO) the
kerosene knockout.  These samples were representative of the headspace
which flowed from the kerosene knockout tank.
     The samples were obtained utilizing the modified EPA Method 110 for
Integrated bag sampling with the evacuated can system and the EPA Method
25 procedure'for Total Gaseous Non-Methane Organic (TGNMO) sampling.
The EKO sample point was 10 inches inside of the port on top of the
Kerosene Knockout Tank.  This point was maintained for monitoring the
temperature with a thermocouple wire secured to the sample probe.   The
flow from this sample location was measured by plant charts of the
orifice reading prior to entry into the knockout trap.   This flow was
assumed to be equal to the flow exiting the atmosphere vent.  The gas
flow from the Atmosphere Outlet vent was inaccessible for measurement.
     The Method 25 sample system was pre-fabricated in the on-site lab
facility and transferred to the lab site.   The sampling procedures were
carried out according to standard procedures outlined in the Guideline
Series 0EPA-450/2-78-041.  The sample was  obtained simultaneously with
the integrated bag sample, packed in dry ice, and shipped to Pollution
Control Science, Inc., of Miamisburg, Ohio for analysis.
5.2.2  Styrene Stripper Systems
     Latex samples were obtained from the  two stripper systems.   The
sample location for the inlet (ESI) was at the stream feeding the stripper.
The sample point was a by-pass valve from  the inlet stream.   The sample
location for the outlet (ESO) was at a line of the stripper outlet used
by the plant for product sampling.   The sample point was maintained at
this sampling valve for both stripper systems.   The location of the
sample points is represented in Figure 2-2.
     The two stripper systems were located in different sections of the
Borger facility.   The inlet and outlet of  each stripper was located in
the same building.  The latex was sampled  at the stream entering and
leaving the stream.   The gas exhaust from  the stripping process was
piped into the exhaust stream that processed through the kerosene absorber,
therefore contained in the EKO sample.   The sample was  obtained from the
inlet and outlet sample points according to the sampling procedure of
                                 5-16

-------
the EPA Draft Method Determination of Residual Styrene in Styrene Butadiene
Latex by Gas Chromatographlc Analysis (Appendix F).   The temperature of
latex was taken directly from the cement sample bottle with a metal
thermometer.  The samples were capped and labeled for storage to be
transported to the TRW lab facility in the Research Triangle Park for
analysis.
     The sampling procedure of the stripper system proceeded when a
blowdown through the stripper was in process.   The latex samples were
obtained at different stages of the blowdown tank period of dumping.
The inlet and outlet samples were taken simultaneously.   No flows could
be taken during the sampling, and since the efficiency of the strippers
was the objective, it was assumed that the inlet and outlet had equal
flows.
5.2.3  Fugitive Roof Vents
     Fugitive emissions from the white and black master production lines
were emitted from roof vents at different locations in the process.  The
sample locations for the roof vents (ERO) are presented in Figure 5-10
and 5-12, with reference to the production process located under the
vent.   Samples were collected through a probe inserted above the fan
which drew emissions from the production building (Figure 5-11 and
5-13).  The sample points were maintained above the fan in each vent by
securing the probe through the side of the vent (see Figure 5-14).  The
velocity of the gas stream exitting the vent was determined with a vane
anemometer.  The measurement points and vent dimensions are presented in
Figure 5-14.  The temperature of the exitting gas stream was monitored
by securing the thermocouple head to the anemometer device.
     The evacuated can system was required for the extraction of the
integrated gas bag sample from the gas stream according to the modified
EPA Method 110.  A trial moisture train run, according to EPA Method 4
procedures, produced a zero moisture determination and this condition
was assumed for all roof vents.
     The sampling procedures were initiated when observations of the
process in the production building confirmed the operation of the finishing
tanks and dryers.  The gas samples were obtained in a series of two
vents at a time.  The velocities were determined during the sampling.
                                 5-17

-------
I
I
L-
r
i
i
            ERO-D
            ERO-C
II
II
n
                UGEND



            I  1  Roof Vents


                 Drier Vents
                                       S'n
                                       L_I~ Vent Ducts
      EDO-A
           ERO-B
            ERO-A
Figure 5-10.   White Line Roof:  Sample Locations

              ERO's and EDO.

-------
   E
                                  ^Drier Vents To Roof
 Dewatering
                        Reslurry
                        Soap Conversion Tank
                        Crumb Solution Inlet
                       Coagulation Tank
                                                   A_. Excess
                                                  \    Ambient
                                                   VJAir
Figure 5-11.  White Line Processing Building.

                       5-19

-------
      ERO-BMA
                             LEGEND



                          .[""] Roof Vents



                              Drier Vents
                          '1 i!
                          !_-_„ Vent Ducts
             Co
             "CUD
                    EDO-tHD
                i-WC
Figure 5-12.  Black Master Roof:  Sample Locations

          ERO-BM's, EDO-BM's.

-------
              Conveyor
                              Drier Vents To Roof
                {  ;")Reslurry

                       Dewatering  Screen
                7 ^
                1   'Soap Conversion Tank


                f J']Coagulation Tank

                         Black  Latex  Feed
                         i
                        Crumb Solution Inlet
0                                                    Excess
                                                    Ambient
                                                    Air
Figure 5-13.  Black Master Processing Building.

                    5-21

-------
The  fan of  roof vent at sample location ERO-BMA was not operating and
therefore,  was not tested.
5.2.4  Dryer Vents
     Gas emissions from the dryer systems were vented to the roof and
into the atmosphere (EDO).  The black master line had eight vents, of
which only  four dryer vents along the system were operating.  Samples
were drawn  from each vent with a 5-foot probe inserted into the roof
stack.  Figure 5-12 presents the locations of the dryer vents and the
locations of the vents along the dryer line.
     The white line had four dryer vents, of which only one was operating.
Samples were taken with a 5-foot probe inserted into the roof stack.  A
pseudo-stack was required to modify the stack to a proper length, enabling
accurate flow measurements according to EPA standards (Figure 5-15).
     The roof stack from the dryers varied in diameter and height.  The
sample point for each was maintained in the center of the stack and at
least 2 feet inside the exit of the stack.  The temperature was monitored
at this point in the gas stream by securing the thermocouple lead to the
sample probe.  The flow measurement points were taken from the various
stacks according to the dimensions of the stack and these points are
presented in Figure 5-16.   The temperature was monitored at these points
by securing the thermocouple lead to the anemometer device.
     The evacuated can system was required for the extraction of the
integrated bag samples from the gas stream according to the modified EPA
Method 110.  A midget impinger train was utilized at the sample points
for  moisture determination according to EPA Method 4.
     The sampling procedures were initiated when observations of the
process in the production building confirmed the operation of the dryers.
The  dryer vents operating and tested are labeled on Figure 5-10 and
5-12.  The vents with no flow were not labeled or tested.   The gas
samples were obtained in a series of two vents at a time.   The velocities
were determined during the sampling period.
                                 5-22

-------
01


Traverse Point
Number

IKS 1
2
3
4
5
6
7
8
9
10


Traverse Point
Centrofd Lo-
cation (From
left and top.
In Inches)
5.4/4.9
16.3/4.9
27.1/4.9
38.0/4.9
. -48.8/4.9
5.4/14.8
16.3/14.8
27.1/14.8
38.0/14.8
48.8/14.8
(Emulsion Roof
Vents)
                     (Traverse Points)

                            Top
          Left
                                                                           Sample ERO
                                                                           Location
                                                                                                          (Side View)
                                                                             Fan
Flow
                                                                                     Probe to sample system from
                                                                                     sample point ERO
                                  3 ft
                                                                                 X/
                                                                                                         (Top View)
Flow
                           Figure 5-14.   Emulsion crumb process roof venturi:  sample points ERO-A,
                                         ERO-B,  ERO-C, ERO-D, ERO-BMB, ERO-BMC, ERO-BMD,

-------
                                                                                            (Side View)
V
r\>
Traverse Point
Nuaber
**8 1
2
3
4
5
6
7
8
9
10
11
12


Traverse Point
Location From
Outside of Port
(Inches)
.4
1.4
2.5
3.6
5.2
7.5
13.5
15.7
17.3
18.5
19.6
20.6
(Cyclone Outlet
Vent)
                                                                                            Pseudo-Stick
                                                                                                 Probe to Sample Systea

                                                                                                 fro* Sample Point COO
                                                                                                  ซ_ (Traverse  Points)^
                                            Figure 5-15.   Emulsion Crumb process, White Line dryer vent:

                                                           Sample point EDO.

-------
ro
            Hood
?f
J D S
Stack
\
H
                                             Sampling Location
Traverse Points
                                                                              Section A-A
Stack
EDO-BMA
EDO-BHB
EDO-BMC*
EDO-BMD
H
66"
72"
72"
78"
D
37.90"
31.62"
24.52"
31.84"
c
29.75"
24.82"
19.25"
25.0"
Traverse Point Locations from Edge of Stack
dl
7.58"
6.32"
8.17"
6.37"
d2
15.16"
12.65"
8.17"
12.74"
d3
7.58"
6.32"
8.17"
6.37"
d4
15.16"
12.65"
8.17"
12.74"
d5
7.58"
6.32"
—
6.37"
d6
15.16"
12.65"
—
12.74"
d7
7.58"
6.32"
—
6.37"
d8
15.16"
12.65"
—
12.74"
                  *0nly four  sampling  locations were  used on  this stack, one In each quadrant.
                       Figure  5-16.   Emulsion crumb process black line dryer vents; sample points
                                     EDO-BMA, EDO-BMB, EDO-BMC, EDO-BMD.

-------
                  6.0  SAMPLING AND ANALYTICAL PROCEDURES

6.1  SOLUTION CRUMB
     Pertinent data compilation and sample collection and analysis provided
information for the characterization of VOC emissions and the efficiency
of the control devices In the solution crumb process.  The process data
monitored were gas flows; measured with an annubar,  a vane anemometer or a
type S Pitot tube, liquid flows; obtained from plant GPM Meters,  tempera-
ture; measured with K type thermocouple with temperature readout  or metal
thermometers, moisture; measured using midget impinger train (EPA Method 3).
Gas samples were collected using the integrated bag  (modified EPA Method 110)
                             <
procedure.  Latex and crumb samples were collected using a latex  grab
collection procedure.  The analytical procedures used for gas analysis
Included gas chromatography with flame ionization detectors (GC/FID) for
hydrocarbon analysis, thermal conductivity detectors (GC/TCD) for fixed
gases analysis and Method 25 procedures from the EPA Guideline Series for
the Measurement of Volatile Organic Compounds (EPA-450/2-78-041).   Cement
and crumb samples were analyzed using the EPA Draft  Method for Residual
Hexanes (Appendix F).
6.1.1  Flash Tank Ammonia Condenser
     Two locations were monitored at the flash tank  ammonia condenser.
The inlet location (SFI) was changed because of in-line process diffi-
culties.  The alternate position was located in a nitrogen or feed line
bleed-off valve of the flash tank accumulator (Figure 5-1).  No method
modifications were required for sampling the inlet location.   The outlet
location (SFO) required sampling modifications because of inaccessability
of the sampling port and the presence of condensate  in the gas flow.

-------
     6.1.1.1  Flow Measurements.  Flow measurements at SFI were not made
because the alternate location In the nitrogen line was unrepresentative
of the actual condenser inlet flow rate.  Therefore, the flow was assumed
to be the same as the flow measured at SFO.  Flow measurements at SFO were
made with a FTL-75 Dieterich Annubar Flowmeter (Figure 5-2).  The location
was modified by building an extended sampling port.  This was necessary so
that proper flow measurement criteria (EPA Methods 1, 2) could be achieved
for annubar measurement and so that the condensate could be removed from
the stream prior to gaseous measurement.  Accuracy of these flow measure-
ments are questionable because the flow had a short duration and was
irregular as well as being affected by the large volume of liquid
solvent in the flow.
     6.1.1.2  Temperature Measurements.   Temperature measurements were
made by securing a thermocouple wire onto the sample probe in the flow at
SFO and monitoring the temperature on the digimite.  (See Figure 5-2.)
Accuracy of this type of temperature measurement is ฑ 2ฐF.
     6.1.1.3  Gas Sample Procedures.  Gas samples were collected at both
SFI and SFO.  The inlet to the ammonia condenser (SFI) sample was taken
from the flash tank accumulator through a nitrogen feed line (Figure 5-1).
An integrated bag sample was collected over the Reactor dump period using
modified EPA Method 110 (Appendix G).   The gas sample at the outlet of the
ammonia condenser (SFO) was taken from the end of the modified sampling
port (Figure 5-2).   The evacuated can method (modified Method 110,
Appendix G) was used to collect the gas sample over the period of high
flow.   There were two significant problems encountered in sampling this
point.   The first was the short duration of the flow.  Flow from the
ammonia condenser occurred only when the flash tank was being filled and
lasted approximately five minutes.  The second problem was the large
volume of condensate (approximately 1 gallon per minute) blown out of the
condenser during the duration of the flow.  This was a non-normal operating
condition, and the modified sampling port allowed the condensate to be
removed from the flow and to be saved for analysis.  The amount of conden-
sate flowing out of the condenser was measured after each test period.
The method of liquid volume quantification was to measure the level of
condensate in the catch tank.   This level was recorded, and a calibration
                                  6-2

-------
  was performed during post-test functions to determine the liquid volume  at
  this level in the catch tank.
       The gas samples from both SFI and SFO were analyzed by GC/FID for
  identification and quantification of hydrocarbons (Appendix F).   Dilutions
  of the samples were required at both Icoations.   (See Dilution Methods,
  Appendix B.)  The samples were also analyzed by GC/TCD for qualification
  on the inert gases present (Appendix F).
       The condensate collected at SFO was analyzed by direct injection  onto
  the column of the GC/FID.   Dilution of the sample was necessary  to be
  within the working range of the instrument (Appendix B).
       6.1.2.4  Moisture Determinations.   Moisture determinations  were not
  made at either SFI or SFO.   The liquid condensate in the flow  at SFO was
  pure organic solvent so the percent of moisture in the gas was assumed to
  be negligible.
       6.1.1.5  Sample Preparation.   The gas samples at SFI and  SFO were
.. analyzed immediately after being sampled.   This was necessary  because
  preliminary studies on the degradation rate of hydrocarbons in the gas
  samples was significant after one hour.   The gas samples were  not pre-
  served because they would be of little value due to degradation.   The
  condensate samples from SFO were preserved in amber glass jars at 4ฐC.
       6.1.1.6  Analytical Procedures.   The analytical  results at  SFI and
  SFO appear satisfactory .   Three sets of integrated bags were  analyzed.
  Dilutions of all  the samples were required and internal  standards were
  used to calculate the dilution ratios (Appendix B).   Repetitive  analyses
  were made on each sample until  two successive repeatable results  (area
  counts) were obtained with 5% (Appendix B).   Tables 3-1 and 3-2,  subsum-
,. mary of emissions at the solution crumb flash tank reports the average of
 , the repeatable values for each hydrocarbon.   The results were  calculated
  in ppm as benzene, since benzene was used to calibrate the instruments.
  Conversion factors (Appendix B) were used to convert the ppm as  benzene
  numbers to ppm as compound numbers.
  6.1.2  Solution Crumb:   Blend Tank Ammonia Condenser
       In the solution crumb process the latex leaves the flash  tank for one
  of three blend tanks.   The vents for the blend tanks  are connected and
                                    6-3

-------
pass through an ammonia condenser before being vented to the atmosphere.
Gas sampling locations were positioned before and after the condenser.
     6.1.2.1  Flow Measurements.  No flow measurements were made at the
inlet to the ammonia condenser (SBI) or the outlet to the ammonia condenser
(SBO) (Figure 5-3).
     6.1.1.2  Temperature Measurement.  The temperature of the gases in
the blend tank (SBI) were measured using a type K thermocouple wire and
Digimite temperature readout.
     6.1.1.3  Gaseous Measurement.  Gas samples were collected using
evacuated can method (modified EPA Method 110).  Since there was no flow
through the ammonia condenser, no sample was collected at the outlet (SBO)
(Figure 5-3).  The inlet sample location was changed since there was no
flow going from the blend tank to the ammonia condenser.   The alternative
location was a port at the top of the blend tank (Figure 5-3).   The inlet
sample was collected by placing a probe into a man-hole cover at the top
of the blend tank and drawing a sample (using modified Method 110) from
the head space in the blend tank.
                              *
     Gas samples from SBI were analyzed by GC/FID for total hydrocarbon
content (Appendix B).  There was one deviation from the method described
in the appendix for gaseous hydrocarbon analyses.   The 20% SP 21007.1%
Carbopack C columns were replaced by a blank column.   The instrument was
calibrated with propane, and the samples were analyzed for total hydrocar-
bons as propane.   Dilutions of the samples were required to lower the
total hydrocarbon content into the working range of the instrument.
     6.1.2.4  Moisture Determinations.  The moisture content of the gas
concentration in the blend tank (SBI) were determined with a midget
impinger train.   The procedure followed was according to the EPA Method 4
for Moisture Determination.
     6.1.2.5  Sample Handling.  The samples taken at the solution crumb
blend tank were analyzed for total hydrocarbons by GC/FID analysis.   The
sample was analyzed immediately after being collected since degradation of
the hydrocarbons will occur.  Significant error in the results were observed
after two hours from collection of sample.  Therefore, the samples were
                                                \
not preserved after being analyzed.
                                  6-4

-------
       6.1.2.6  Results.  The results appear questionable because of the
  flow to the sample system.  Section 3 presents an explanation at this
  sample location.
  6.1.3  Solution Crumb;  Steam Stripper
       No gaseous emissions are vented from the steam stripper.  At this
  point' in the process, the latex cement Is converted to latex crumb by
  treating the cement with stream.   Samples of the latex as cement and crumb
  were collected and analyzed at this location.
       6.1.3.1  Flow Measurement.  Flows of the cement into the steam stripper
  are monitored by plant personnel  and, therefore, flow measurements for
  this test were taken from plant flow meters.
       6.1.3.2  Temperature Measurements.   The temperature from the Latex
  sample obtained at the sample location was determined by a direct readout
  from a metal thermometer placed in the sample jar during sampling.
;      6.1.3.3  Latex Measurements.   At the inlet to the steam stripper  the
'" latex is in the cement3 phase.
       The latex grab collection procedure (Appendix G) was used to obtain
  the inlet sample.  An added precaution was taken with the cement samples,
  in order to minimze loss of cyclohexane, the collection jar was overflowed
  and capped leaving no air space in the jar.   The sample was stored at  4 C
  to further reduce the possibility of loss.   To another jar, known weight
  of cement was added to 200 ml methylene chloride for GC/FID analysis (EPA
  Method for Determination of Residual Hexanes,  Appendix F).
       In the steam stripper, steam is forced through the cement, which
  converts the cement into a crumb.   A grab sample of crumb was collected as
  described in the EPA Draft method for Determination of Residual Hexanes
; (Appendix F).  A known weight portion of the crumb (approximately 10 g)
  was added to 200 ml of methylene chloride as described in the EPA draft
  method (Appendix B) for GC/FID analysis.  The remaining crumb was stored
  in an inert plastic, air tight bag at 4ฐC.
       Since the latex samples were totally soluble in methylene chloride.
  Analysis of both the cement and crumb samples performed by direct liquid
  aPlant nomenclature for solution crumb latex.   In this phase the latex
   1s a viscous liquid of primarily cyclohexane.
                                    6-5

-------
 Injection or a GC/FID (Appendix F).  The concentrations were calculated
 based on the known weight of sample added to the methylene chloride
 (Appendix B).
     6.1.3.4  Dry Weight Determinations.  Known weights of both the cement
 samples (SSI) and crumb samples (SSO) were dried at 105ฐC until two conse-
 cutive weightings were repeatable within .0005 g.  For the cement samples,
 this method is unacceptable.  At this point in the process the latex
 cement is composed primarily of cyclohexane; functioning as a transport
 agent.  Since the flashpoint is lower than 105ฐC both the cyclohexane and
 water will be removed from the sample.  Therefore, the hexane concentration,
 when reported on a gram per gram dry basis shows higher than 100% reading
 for hexanes (Appendix B).
     At the steam stripper outlet, the crumb has been treated with steam.
 This process will effectively remove a majority of the cyclohexane and dry
 weights then, will represent percent moisture loss (Appendix B).
     6.1.3.5  Results.  Results of process measurements were the average
 of the readings taken over the course of the test.  The field process data
 for each of the three tests conducted at SSI and SSO appear in Appendix D.
 The analytical results appear in the sub-summary as the average of each
 test.  The analytical results, reported as peak areas must be repeatable
within 5%.   Sample calculations (Appendix B) show that the peak area is
 converted to a concentration based on a K-factor of the standards and how
 the dry weights and dilution factors are used to arrive at the sample
concentration.
6.1.4  Solution Crumb:  Crumb Tank
     The crumb coming out of the steam stripper flows into the crumb tank
 using water as the transport medium.  The gas emissions out of this tank
are vented to the atmosphere.   Integrated bag samples were collected at
the outlet of the solution crumb tank.
     6.1.4.1  Flow Measurements.   Flow measurements were performed with a
vane annemometer.  The velocity measured was from the center of the stack
and successive runs were performed to obtain an average flow.
     6.1.4.2  Temperature Measurements.   Temperature readings were taken
during the sampling period by means of a type K thermocouple wire attached
to a Dignimite temperature display.
                                  6-6

-------
   - 6.1.4.3  Gaseous Measurement.  Integrated bag samples were taken from
the emission flow above the solution crumb taken by the modified EPA Method
110 (Appendix G).
     6.1.4.4  Sample Handling.  Preparation of the gas samples for hydro-
carbon analysis included dilution of the sample to within working range of
the instrument (Appendix B).  Internal standards were used to calculate
the dilution ratio (Appendix B).  Analyses of the samples were performed
by GC/FID for hydrocarbons (Appendix F) and by GC/TCD for inert gases
(Appendix F).
     6.1.4.5  Moisture Determination.   The moisture content of the gas
concentration exiting the solution crumb tank (SCO) was determined with a
midget impinger train.  The procedure followed was according to the EPA
Method 4 for Moisture Determination.
     6.1.4.6  Sampling Handling.  The gas samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to be a
significant error after two hours (see Section 6.4).   For the same reason,
samples were not preserved.
     6.1.4.7  Results.  Results of process measurements appear in the
summary (Table 2-2) as are averages of the readings taken over the course
of the test.  The field process data for each of the three tests conducted
at SCO appear in Appendix D.  The analytical results appear in the sub-
summary as the average of each test.  Analytical results reported as peak
area must be repeatable within 5% for a test.  Peak areas are converted to
concentrations by using a constant derived from calibration standards
(Appendix A).  Any unidentified compounds were classified as the identi-
fied compound they most closely resembled in terms of retention time.
Therefore, some summary and sub-summary results include these unidentified
compounds (see computer sheets, Appendix A).  The concentrations were then
multiplied by the calculated dilution ratio (Appendix B and converted to
dry weight basis concentrations.
6.1.5  Dewatering Screens
     The latex crumb uses water as a transport medium which must be
removed to prepare the final product.   The crumb flows from the crumb tank
onto the dewatering screens.  The water effluent is recycled to the steam
                                  6-7

-------
stripper.  A hood over the screens captures the gaseous emissions which
are vented on the roof.  Gas samples were taken at the vent while the
screens were operating.
     6.1.5.1  Flow Measurements.  The outlet flow of the roof vent ducted
from the dewatering screen hood was measured according to EPA Method 1.
The vent exit was modified with a pseudo-stack to adapt the vent according
to the EPA criteria for flow measurments with a S-type pitot tube.
     6.1.5.2  Temperature Measurement.  Temperature readings were taken
using a type K thermocouple wire attached to a Digimite temperature display.
     6.1.5.3  Gaseous Measurement.  Integrated bag samples were taken from
the roof vent over the dewatering screens while they were operating following
the modified EPA Method 110 (Appendix G).
     6.1.5.4  Moisture Determinations.  The moisture content of the gas
concentration exiting the roof vent from the dewatering screen hood was
determined with a midget impinger train.  The procedures followed were
according to the EPA Method 4 for Moisture Determination.
     6.1.5.5  Sample Handling.  The samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to be a
significnat source of error after two hours (see Section 6.4).   For the
same reason, the samples were not preserved.
     6.1.5.6  Results.  Results of the process measurements appear in the
summary (Table 2-2) as an average of the readings taken over the course of
the test.  The field process data for each fo the three tests conducted at
SHO appear in Appendix D.   The analytical results appear in the sub-summary
as the average of each test.  The final average does not include the
analytical results from SHO-2 because the sample bag was nearly empty when
received, and the results are unreliable at best.  Acceptable analytical
results, reported as peak area, must be repeatable within 5% for a test
(Appendix B).   Peak areas were converted to concentrations by using a
constant, derived from calibration standards (Appendix B).   These concen-
trations are then converted, by a response factor, to concentrations as
compounded on a dry weight basis (Appendix B).   Any unidentified compounds
were classified as the identified compound they most resembled, in terms
of retention time.  Therefore, some summary and sub-summary results include
these unidentified compounds (see computer sheets, Appendix A).
                                  6-8

-------
6.1.6  Solution Crumb Cyclone
     In the solution crumb process the crumb form exits the pelletizer for
the finishing drier.  The crumb passes through a cyclone device.   The
crumb is suspended in the air current to allow for the heavier crumb in
pellet form to drop out onto the conveyor for transport into the drier.
While the gas emission remaining from the process is captured and emitted
with the cyclone air current.  A sampling location was positioned at the
roof vent exit from the cyclone device.
     6.1.6.1  Flow Measurements.  The outlet flow of the roof vent ducted
from the cyclone separating device were measured according to EPA Method 1.
The vent exit was modified with a pseudo-stack to adapt the vent according
to the EPA criteria for flow measurements with a S-type pilot tube.
     6.1.6.2  Temperature Measurement.  Temperature readings were taken
using a type K thermocouple wire attached to a Digimite temperature display.
     6.1.6.3  Gaseous Measurement.  Integrated bag samples were taken from
the roof vent of the cyclone while the process was operating.   Samples
were taken following the modified EPA Method 110 (Appendix G).
     Analysis of the samples was performed by GC/FIO for hydrocarbons
(Appendix F) and by GC/TCD for inert gases (Appendix F).  No dilutions
were required prior to analysis and as modifications of the method were
necessary.
     6.1.6.4  Moisture Determinations.  The moisture content of the gas
concentration exiting the roof vent from the cyclone separating device was
determined with a midget impihger train.  The procedure followed was
according to the EPA Method 4 for Moisture Determination.
     6.1.6.5  Sample Handling.  The samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to be a
significant source of error after two hours (see Section 6.4).  For the
same reasons the samples were not preserved.
     6.1.6.6  Results.  Results of the process measurements appear in the
summary (Table 2-2) as an average of the field process data taken over the
course of the test (Appendix B).
     The analytical results appear in the sub-summary as the average of
each test.  All three test runs were included in the average in the summary.
                                  6-9

-------
 Acceptable  analytical  results,  reported as peak area, must be  repeatable
 within  5% for  a  test  (Appendix  B).   Peak areas were converted  concentrations
 by using  a  constant,  derived  from calibration standards (Appendix B).
 These concentrations  are then converted, by a response factor, to concen-
 trations  as compounded on a dry weight basis (Appendix B).  Any unidentified
 compounds were classifed as the identified compound they most  resembled,
 in terms  of retention  time.   Therefore, some summary and sub-summary
 results include  these  unidentified compounds (see computer sheets,
 Appendix  A).
 6.1.7  Drier Vents
     There  were  five  identical  drier vents associated with the solution
 crumb process.   The sampling  and analytical techniques applied to each
 drier was the  same, therefore,  the following descriptions are applicable
 to all five.   The analytical  results differ because of the location of any
 given vent  above the process  (see Section 5).  Three sets of integrated
 bag samples were collected from each of the five vents.
     6.1.7.1   Flow Measurements.  The outlet flow of the roof vent ducts,
 from the  varied  locations along the drier were measured according to EPA
 Method 1.  The vent exit did  not conform to the criteria for standard flow
 measurements.   Therefore, sample ports were installed in the roof vent
 ducting.  The  flow measurements were taken from these ports with the
 S-type pitot tube.
     6.1.7.2  Temperature Measurement.  Temperature readings were taken
 during the sampling period using a type K thermocouple wire attached to a
 Digimite  temperature display.
     6.1.7.3  Gaseous Measurements.   Integrated bag samples were taken
 from the  emission flow from the vents above the solution crumb drier using
 the modified EPA Method 110 (Appendix G).   Analysis of the samples was
performed by GC/FID for hydrocarbons (Appendix F) by GC/TCD for inert
gases (Appendix  F).  No dilutions were required prior to analysis.
     6.1.7.4  Moisture Determinations.  The moisture content of the gas
                                               %
concentration exiting the roof vent from the drier outlets were determined
with a midget  impinger train.   The procedure followed was according to the
 EPA Method 4 for Moisture Determination.
                                  6-10

-------
    - 6.1.7.5  -Sample Handling.  The gas samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to be a
significant error after two hours (see Section 6.4).  For the same reason,
samples were not preserved.
     6.1.7.6  Results.  Results of the process measurements appear in the
sub-summary (Tables 3-6, 3-7, 3-8, 3-9, 3-10) as an average of the field
process data measured over the course of the test (Appendix D).  Any
unidentified compounds were classified as the identified compound they
most resembled, in terms of retention time.  Therefore, some summary and
sub-summary results include these unidentified compounds (see computer
sheets, Appendix A).  Notice that the third test SDO-3 at each vent has
higher emissions due to the difference in operating temperature of the
drier as discussed in Section 3.  All three test runs are included in the
average for the summary (Table 2-3).  Acceptable analytical results,
reported as peak area, must be repeatable within 5% for a test (Appendix B).
Peak areas were converted to concentrations by using a constant, derived
from calibration standards (Appendix B).   These concentrations are then
converted, by a response factor, to concentrations as compounded on a dry
weight basis (Appendix B).
6.2  EMULSION CRUMB PROCESS
     The second process line studied at the Borger facility was used in
the formation of styrene-butadiene rubber.  Data and sample collection
provided information for the characterization of VOC emissions and the
efficiency of the control devices.  The process splits into two product
lines after the steam strippers.  The two product lines are referred to as
the white line and the black master line.  The first three points samples
are common to both product lines.   The remaining sample locations are
analogous points from each different product line (see Section 5).  The
process data monitored gas flows measured with either a vane anemometer,
or a type S pilot tube.  Also monitored were liquid flows, obtained from
plant GPM meters, temperature, measured wth K type thermocouple with
temperature readout, and moistures, measured using midget impinger trains
(EPA Method 4).  Gas samples were collected using the integrated bag
procedure (modified EPA Method 110) or using EPA Method 25 procedures for
                                  6-11

-------
TGNM.   Latex samples were collected using a latex grab collection procedure.
The analytical procedures for gas analysis included GC/FID for hydrocarbons,
GC/TCD  for inert gases, and gas analysis for TGNMO (Method 25).  The
analytical procedure for latex analysis is described in the EPA Draft
Method  for Residual Styrene (Appendix F).
6.2.1   Kerosine Absorber Outlet
     The Kerosine absorber outlet is the first point where emissions are
vented  from the emulsion process.  The gaseous emissions from the flash
tank (see process description, Section 4) and from the steam stripper pass
through a chiller, the kerosine absorber and finally a kerosine knockout
before  being vented to the atmosphere.  Gas samples were collected after
the knockout.
     6.2.1.1  Flow Measurements.  The flow into the knockout tank was
monitored by the plant with an orifice in line of the chiller and kerosine
knockout tank to be recorded a square root chart.  The flow into the tank
was seen representative of the outlet vent from this closed system.   The
square  root chart was provided'by Phillips Petroleum with the process
information.
     6.2.1.2  Temperature Measurements.  Temperature measurements were
made during the sampling period using a type-K thermocouple wire attached
to a Oigimite temperature display.
     6.2.1.3  Gaseous Measurements.   Integrated bag samples were taken
from the emission flow from the vents above the solution crumb drier using
the modified EPA Method 110 (Appendix G).  Gas samples for TGNMO analysis
(EPA Method 25) were simultaneously taken at this location.  Gas analysis
by EPA Method 25 was performed by Pollution Control Science, Inc.   FID
analysis for total hydrocarbons was performed in the field (Appendix B).
A modification to the method was required because of the high hydrocarbon
content.  The sample was diluted and injected into a blank column.   The
results of this type of analysis are total hydrocarbons.   There is no
separation of compounds.   The gas samples were also analyzed by GC/TCD for
inert gases (Appendix F).   Acceptable analytical results, reported as peak
area, must be repeatable within 5% for a test (Appendix B).  Sample ERO-BMB-1
was nearly empty so the analytical results used to calculate the hydrocarbon
                                  6-12

-------
concentrations have a weak base.  Peak areas were converted to concentrations
by using a constant, derived from calibration standards (Appendix B).   Any
unidentified compounds were classified as the identified compound they
nost closely resembled in terms of retention time.  Therefore, some summary
and sub-summary results include these unidentified compounds (see computer,
sheets, Appendix A).  The concentrations, as they appear on the summaries,
have been converted to dry weight basis concentrations (sample calculation,
Table A).
     6.2.1.4  Moisture Determinations.  A moisture determination from the
EKO sample location was not required because of the plant providing the
gas flow data.  The gas content of the stream is assumed to be zero around
a control system utilizing kerosine as the absorbent.
     6.2.1.5  Sample Handling.   The gas samples were analyzed by FID
immediately after sampling because degradation of the hydrocarbons was
found to be a significant error after two hours (see Section 6.4).  Samples
collected for TGNMO analysis were stored in dry ice and shipped to Pollution
Control Science, Inc.   FID samples were not preserved after analysis.
     6.2.1.6  Results.   Results of the process measurements appear in the
sub-summary (Table 3-11) as an average of the field process data measured
over the course of the test (Appendix D).  The TGNMO results are listed in
the Appendix B and in Table 6-1 which presents the TGNMO results with the
FID results.   The samples analyzed by FID were diluted prior to analysis.
The dilution factor was calculated based on GC/TCD analysis of both the
diluted and undiluted samples.   One sample's (EKO-1) FID results are based
on one sample injection.  The other results appear as the average of two
repeatable (within 5%) analyses (Appendix B).  Peak areas were converted
to concentrations by using the calibration constant for propane (Appendix A).
These concentrations were multiplied by the dilution factor to give a
final concentration.
6.2.2  Emulsion Crumb Steam Stripper Inlet
     Latex flows from the blend tank to one of three steam strippers.   Two
of the steam strippers were analyzed (B and E) because A and B were assumed
to be the same.
                                  6-13

-------
                       Table 6-1.  TGNMO AND FID RESULTS
                               TGNMO                         FID
     Sample                 (ppm as Cj)                   (ppm as C3)

     EKO-1                    109,251                        88,590a
     EKO-2                     39,021                        65,244
     EKO-3                     52,752                       111,421

a6ased on one sample injection.

     The gaseous emissions vented from the steam stripper are combined
with the emissions from the flash tank (see Section 6.2.1).   This section
deals with sampling analyses of the latex flowing into the steam stripper.
     6.2.2.1  Flow Measurement's.   Flows into the stripper systems were
monitored by plant personnel.  The latex production rate (gallons per
minute) through the stripper systems was provided by Phillips Petroleum.
     6.2.2.2  Temperature Measurements.  The temperature of the latex
sample obtained at the sample location was determined by a direct readout
from a metal thermometer placed in the sample jar during sampling.
     6.2.2.3  Latex Measurements.   At the inlet to the steam stripper the
integrated latex grab method (Appendix G) was used.   Preparation of sample
for residual styrene analysis consisted of adding an aliquot of sample to
methylene chloride containing TBC.   TBC acts as a stabilizer for the
residual styrene.  Samples were stored at 4ฐC.   The latex in methylene
chloride samples were analyzed by direct liquid injection into the GC/FID
according to the EPA Draft Method:   Determination of Residual Styrene
(Appendix F).  The method calls for dissolution of an aliquot of sample in
methylene chloride containing TBC.   Some of the samples were insoluble in
methylene chloride.  These samples were extracted three times with portions
of methylene chloride containing TBC (see Section 6.3).  Dilutions or
extraction corrections were calculated based on the volume of sample and
volume of methylene chloride (Appendix B).
                                  6-14

-------
     6.2.2.4  Dry Weight Determinations.  Known weights of the latex
samples were dried at 105ฐC until two consecutive weightings were repeatable
within .OOOSg.
     6.2.2.5  Results.  Results of the process measurements appear in the
sub-summary (Table 2-5) as the average of the readings taken over the
course of the test.  The field process data for each of the three tests
conducted at the two steam stripper inlets appear in Appendix D.
     The analytical results appear in the sub-summary as the average of
field tests performed on each run.  The field results, reported as peak
areas must be repeatable within 5%.  Sample calculations show how the peak
area is converted to concentration by multiplication of the area and the
calibration factor.  Calibration factors (Appendix B) were calculated on a
daily basis.  The dilution factors were determined by the extracted or
dissolved volumes (sample calculations, Table A).  The dry weights and
dilution factors are applied to the concentration to arrive at the sample
concentration in the sub-summaries (Appendix B).
6.2.3  Emulsion Crumb:  Steam Stripper Outlet
     Steam is forced through the latex in the steam stripper to remove
residual styrene.  The gaseous emissions are vented through the chiller,
kerosine absorber, and kefosine knockout.  After steaming, the latex flows
out of the stripper into storage tanks.  Process data and analytical
measurements were made on the latex at the steam stripper outlet.  The
Borger facility has three steam strippers.   For this test, two were analyzed
                   *      *
(B and E) because A  and B  were assumed to be the same.
     6.2.3.1  Flow Measurements.  Outlet flows from the stripper systems
were monitored by plant personnel.  The latex production rate (gallons per
minute) through the stripper systems was provided by plant personnel.
     6.2.3.2  Temperature Measurement.  The latex sample temperature
obtained at the sample location was determined by a direct readout from a
metal thermometer placed in the sample container during sampling.
     6.2.3.3  Latex Measurement.  The latex samples were collected by the
latex grab method (Appendix G).  Briefly, the method calls for the collection
*
 Plant designations.
                                  6-15

-------
of a  representative integrated latex sample.  A pure sample of latex was
collected for dry weight determination.  This sample was capped and stored
at 4ฐC.  This sample was prepared for GC/FID analysis.  This method required
a known volume of sample be added to 200 ml methylene chloride containing
TBC (EPA Draft Method:  Determination of Residual Styrene).  The TBC
serves to stabilize the residual styrene.  The outlet samples were insoluble
in methylene chloride.  So they were extracted three times with 100 ml of
methylene chloride (Appendix G).  Both the extracted latex and extract
were  stored at 4ฐC.  Extraction corrections were calculated based on the
volume of sample and volume of methylene chloride.
      6.2.3.4  Dry Weight Determinations.  Known weights of the latex
samples were dried at 105ฐC until two consecutive weightings were repeatable
within .OOOSg.  The dry weights appear in Appendix B.
      6.2.3.5  Results.  Results of the process measurements appear in the
sub-summaries (Table 2-5) as the average of the readings taken over the
course of the test.  The field process data for each of the three tests
conducted at the two stripper outlets appear in Appendix D.  The field
                             i
results, reported as peak areas must be repeatable within 5%.
      Sample calculations show how the peak are is converted to concentration
by multiplication of the area and the calibration factor.  Calibration
factors (Appendix B) were calculated on a daily basis.   Dilution factors
were  determined by the volumes used in the extractions or dissolved samples.
The dry weights and dilution factors are applied to the concentrations
(Appendix B) to arrive at the concentrations reported in the sub-summaries.
6.2.4  Emulsion Crumb:  Process Split
     After steam stripping the latex flows into a blend tank where additives
are put into the latex which will determine the final product.   Two products
are produced:   a white latex rubber through the white line process and a
black latex rubber through the black master line process.  The processes
follow the same basic production steps (see Process Description, Section 4)
only  the additives differ.   Therefore, the sampling and analytical methods
describe, for a location, apply to the same location for both the black
master and white lines.
                                  6-16

-------
6.2.5  Emulsion Crumb:  Roof Vent A
     After leaving the blend tank, the latex is treated with water,  soap,
and other additives in a series of tanks.   The latex flows onto a screen
allowing the water falls out.  A roof vent above the screens vented  the
process emissions to the atmosphere.   Process data and samples for analysis
were collected at this vent.
     6.2.5.1  Flow Measurements.  The outlet flows from the roof vents
were fan forced out a two-sided vent.  The outlet area from both sides of
the vent were identical.  A vane anemometer was utilized to measure  the
velocity at standard points based on the outlet area.   The measured  flows
from the two sides were combined to give the total flow from the vent.
     6.2.5.2  Temperature Measurements.   Temperature readings were taken
during the sampling period by means of a type K thermocouple wire attached
to a Digimite temperature display.
     6.2.5.3  Gaseous Measurement.  Integrated bag samples were collected
from roof vent A of the white line (ERO-A).   No sample was collected from
an analogous vent in the black master line (ERO-BMA),  since this vent was
inoperative during the sampling.  Modified EPA Method 110 was followed in
sample collection (Appendix G).
     Analysis of the samples was performed by GC/FID for hydrocarbon
content (Appendix F) and by GC/TCD for inert gases (Appendix F).  No
dilutions were required prior to analysis.
     6.2.5.4  Moisture Determination.  The moisture content of the
fugitive emissions from the emulsion finishing system was assumed to be
zero.
     6.2.5.5  Sample Handling.  The samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to  be a
significant source of error after two hours (see Section 6.4).  For  the
same reason, the samples were not preserved.
     6.2.5.6  Results.  Results of the process measurements for roof
vent A of the white line (roof vent A of the black master was inoperable)
appear in the sub-summary (Table 3-12) as an average of the field process
data taken over the course of the test (Appendix D).
                                  6-17

-------
     The analytical results appear In the sub-summary as the average of
each field test.  All three samples were included in the average in the
summary.  Acceptable analytical results, reported as peak area, must be
repeatable within 5% for a test (Appendix B).  Samples analyzed from
Runs 1 and 2 (ERO-A-1 and ERO-A-2) do not meet this criteria.  Only one
analytical run was performed on each of these samples.  Peak areas were
converted to concentrations by using a constant, derived from calibration
standards (Appendix 8).  These concentrations were then converted, by a
response factor, to concentrations as compounded on a dry weight basis.
6.2.6  Emulsion Crumb:  Roof Vent B
     Roof vent B from the white line process (ERO-B) and roof vent B from
the black line process (ERO-BMB) were both located above the presser of
each process.  Process data and samples for analysis were collected at
this location.
     6.2.6.1  Flow Measurements.  The outlet flows from the roof vents
were fan forced out a two-sided vent.  The outlet area from both sides
were identical.  A vane anemometer was utilized to measure the velocity at
standard points based on the outlet area.   The measured flows were combined
as the total flow from the vent.
     6.2.6.2  Temperature Meaurements.  Temperature readings were taken
during the sampling period by means of a type K thermocouple wire attached
to a Digimite temperature display.
     6.2.6.3  Gaseous Measurements.  Integrated bag samples were collected
from roof vent B of the white line (ERO-B) and from an analogous vent in
the black master line (ERO-BMB).  Modified EPA Method 110 was followed in
the collection of the sample (Appendix G).
     Analysis of the samples was performed by GC/FID for hydrocarbon
content (Appendix F) and by GC/TCO for inert gases (Appendix F).  No
dilutions were required.
     6.2.6.4  Moisture Determination.  The moisture content of the fugitive
emissions from the emulsion finishing system was assumed to be zero.
     6.2.6.5  Sample Handling.   The samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to intro-
duce significant error after two hours (see Section 6.4).   For the same
reason samples were not preserved.
                                  6-18

-------
       6.2.6.6  Results.   Results  of the process  measurements  for  roof  vent
  B of the white line (ERO-B)  and  of the black master  line  (ERO-BMB)  appear
  in sub-summary Tables 3-13 and 3-19 as an  average  of the  field process
  data taken over the course of the test (Appendix 0).
       The analytical results  appear in the  sub-summary as  the average  of
  each field test.
  6.2.7  Emulsion Crumb:   Roof Vents C & D
       The roof vents C and 0  from the white line and  black line processes
  (ERO-C and ERO-D and ERO-BMC and ERO-BMD,  respectively) were the same and
  were all located above the driers in each  process.   Process  data and
  samples for analysis were collected at this location.
       6.2.7.1  Flow Measurements.   The outlet flows from the  roof vents
  were fan forced out a two-sided  vent.   The outlet  area from  both sides of
  the vent were identical.   A  vane anemometer was utilized  to  measure the
i- velocity at standard points  based on the outlet area.  The measured flows
v from both sides of each vent were combined to give the total  flow from
 i that vent.
i      6.2.7.2  Temperature Measurements.  Temperature readings were  taken
  during the sampling period by means of a type K thermocouple wire attached
  to a Oigimite temperature display.
       6.2.7.3  Gaseous Measurements.   Integrated bag  samples  were collected
  from ERO-C, ERO-D, ERO-BMC,  and  ERO-BMD following  the modified EPA  Draft
  Method 110 (Appendix F).
       Analysis of the samples was performed by GC/FID for  hydrocarbon
  content (Appendix F) and by  GC/TCD for inert gases (Appendix F).  No
  dilutions were required.
 <      6.2.7.4  Moisture Determinations.   The moisture content of  the
  fugitive emissions were assumed  to be zero.
       6.2.7.5  Sample Handling.   The samples were analyzed immediately
  after sampling because degradation of the  hydrocarbons was found to intro-
  duce significant error after two hours (see Section  6.4).  For the  same
  reason samples were not preserved.
       6.2.7.6  Results.   Results  of the process  measurements  for  ERO-C,
  ERO-D, ERO-BMC, and ERO-BMD  appear in the  sub-summary (Tables 3-14, 3-15,
                                    6-19

-------
3-18, 3-19) as an average of the field process data taken over the course
of the test (Appendix D).
     The analytical results appear in the sub-summary as the average of
each field run.  Acceptable analytical results, reported as peak area,
must be repeatable within 5% for a test (Appendix B).  Peak areas were
converted to concentrations by using a constant, derived from calibration
standards (Appendix B).  Any unidentified compounds were classified as the
identified compound they most closely resembled in terms of retention
time.  Therefore, some summary and sub-summary results include these
unidentified compounds (see computer sheets, Appendix A).  The concen-
trations, as they appear in the summaries have been converted to dry
weight basis concentrations.
6.2.8  Drier Vents
     The final process step prior to pressing and packaging is the drying
of the pelletized rubber.  There were two driers associated with the white
line and two driers associated with the black master line.   Four outlets
for each drier were vented through four dryer vent stacks located on the
roof (see Figures 5-10 and 5-12).  In the white line process only one
drier was operating and only one of the four vents was operating.   In the
black master line process both driers were operating and all four vents
were operating.
     6.2.8.1  Flow Measurements.  The flows out of the black master line
drier stacks were measured with a vane anemometer.  The dimensions of the
stacks varied, therefore, different points were measured according to the
outlet area of the stack (Section 5).  The measured velocity flows were
averaged.
     The flow determination out of the white line drier stack required a
pseudo-stack.   The pseudo-stack adapted the stack exit to conform with the
EPA criteria for flow measurements with a S-type pitot tube according to
EPA Method 1.
     6.2.8.2  Temperature Measurements.   Temperature readings were taken
using a type K thermocouple were attached to a Digimite temperature display.
     6.2.8.3  Gaseous Measurements.   Integrated bag samples taken from the
only operational  white line drier vent (EDO-1, 2 and 3) and from the black
                                  6-20

-------
naster line drier vents (EDO-BMA, EDO-BMB, EDO-BMC and EDO-BMD) following
the modified EPA Method 110 (Appendix G).
     Analysis of the samples was performed by GC/FID for hydrocarbons
(Appendix F) and by GC/TCD for inert gases (Appendix F).  No dilutions
were required.
     6.2.8.4  Moisture Determinations.  The moisture content of the gas
concentration exiting the driers were determined with a midget impinger
train at the sample point.  The procedure followed was according to the
EPA Method 4 for moisture determination.
     6.2.8.5  Sample Handling.  The samples were analyzed immediately
after sampling because degradation of the hydrocarbons was found to introduce
significant error after two hours (see Section 6.4).  For the same reason,
samples were not preserved.
     6.2.8.6  Results.  Results of the process measurements for the drier
vents appear in the sub-summaries as an average of the field process data
taken over the course of the test.
     The analytical results appear in the sub-summary as the average of
each field test.  Acceptable analytical results, reported as peak area,
must be repeatable within 5% for a test (Appendix B).  Peak areas were
converted to concentrations by using a constant, derived from calibration
standards (Appendix B).  Any unidentified compounds were classified as the
identified compound they most closely resembled in terms of retention
time.  Therefore, some summary and sub-summary results include these
unidentified compounds (see computer sheets, Appendix A).  The concentrations
as they appear in the summaries have been converted to dry weight basis
concentrations.
                                  6-21

-------
                          7.0  METHOD DEVELOPMENT

7.1  SAMPLING
     The sampling methods developed for this test include the grab sampling
method for the latex cement and latex crumb, the modified moisture trains
used at the solution crumb ammonia condenser, the integrated, explosion
proof can method used for gas sampling, and anemometer flow meaurements.
7.1.1  Grab Sampling Method
     Integrated grab samples were collected at the inlet and outlet of the
steam strippers in both the solution crumb process and the emulsion crumb
process.   The basic grab sampling method calls for collection of an inte-
grated sample during process operation (Appendix F).   Care is taken to
assure that contamination or loss of sample is avoided.   For example,  all
the samples are stored at 4ฐC in airtight glass containers.   Specific
precautions are defined by the conditions at a particular location.  In
the solution crumb process, the inlet latex sample required special  handling.
The cement uses an organic solvent as the transport medium and loss of
hydrocarbons through evaporation could introduce significant error during
sampling.  To avoid this problem a latex sample was collected by allowing
the cement to overflow the collection jar such that when capped, there is
no head space in the jar.  A second sample was collected by measuring  the
weight added to a known volume of methylene chloride so that no hydrocarbons
are lost by measuring a known volume.
     The crumb collected at the outlet of the steam stripper was collected
using the method described,in the EPA Draft Method - Determination of
Residual  Hexanes (Appendix F).  In brief, the method calls for collection
of sample by allowing the port to flush for several seconds, catching  the
crumb in a strainer so that the crumb is retained, but the water falls out
and then taking a representative sample from the collected crumb.   The
crumb samples were stored in air tight, inert-plastic bags at 4 C.

-------
     The samples collected at both the inlet and the outlet of the
emulsion crumb steam stripper followed the method described in the EPA
draft method - Determination of Residual Styrene (Appendix F).  In brief,
the method calls for adding a portion of the styrene latex, from the
process, to methylene chloride containing TBC.  TBC stabilizes the styrene
so that no styrene is lost through continued polymerization.
7.1.2  Moisture Trains
     The moisture determination procedure for gas streams were according
to the standard EPA Method 4 from the Federal Register.
7.1.3  Integrated Explosion Proof Can
     The explosion proof sampling system was designed by the TRW sample
team to obtain an integrated bag system.  The evacuated can provides the
sampling capability as explained in Appendix G.
7.1.4  Anenometer Flow Meausurements
     The anemometer measurements utilizes the theories of the S-type pitot
tube measurements.   The traverse points are positioned at the same coordi-
nates.  The anemometer is used primary for the stack with low flows that
do not meet the requirements of a pitot tube measurement by EPA Method 1
or 2.
7.2  ANALYTICAL METHODS
     The analytical methods developed for this test included procedures
for hydrocarbon analysis on the three latex forms and on the gaseous
emissions.   The three latex forms were the solution process cement and
crumb and emulsion process cement.
7.2.1  Latex Analysis
     The latex analysis methods followed were from the EPA Draft methods
for the Determination of Residual Hexanes (Appendix F) and the Determination
of Residual Styrene (Appendix F).  All hydrocarbon analyses were conducted
by direct liquid injections on a GC/FID system (Figure F).   In the solution
crumb process two types of latex were collected, cement and crumb.   The
cement and crumb samples collected were both soluble in methylene chloride.
A known weight of sample was dissolved in methylene chloride and injected
into a 12'  by 1/8"  stainless steel  column packed with 20% SP2100/.1*
Carbopack C at 100ฐC.  All latex samples were agitated on an Eberbach
                                  7-2

-------
wrist-action shaker for one hour prior to injection.  Sample injection
size was 1 ml with a 1 ml methylene chloride flush.   The Shimadzu GC/FID's
use a differential mode for analysis (Figure F-2).  The detection limit of
the instruments is approximately 0.1 ppm.
     In the emulsion crumb process the styrene latex was a problem.   Some
of the samples were soluble in methylene chloride, others were not
(Appendix B).  The insoluble samples were extracted with methylene chloride
containing TBC.  An internal standard, . 1% undecane by volume, was added.
A study on the removal efficiency of the extractions was performed
(Table F-l).  Three extractions were found to remove greater than 90% of
the residual styrene in the samples.  For the insoluble latex samples a
50 ml aliquot of sample was extracted three times with 100 ml portions of
methylene chloride.  Each extraction was agitated for four hours in an
Eberbach shaker and then separated.   For the latex samples which were
soluble in methylene chloride, a 50 ml aliquot of sample was dissolved in
200 ml methylene chloride.
     Analyses were performed pn Shimadzu Mini 2's.  For the analysis of
the styrene latex, the GC/FID-was equipped with two 6' by 1/8" stainless
steel columns packed with 1.75% at 1200/Bentone 34.   The dual columns were
used because the Shimadzus work in a differential mode.   Column tempera-
ture was set at 100 C.  Direct liquid injection onto the column of 1 ml of
sample with a 1 ml solvent flush was used.  The measureable styrene limit
appears to be . 1 ppm on the scale used for these analyses.
     Degradation studies of styrene latex in methylene chloride and
cyclohexane latex in methylene chloride showed that samples stored at 4ฐC
remained stable for a period of two weeks.
7.2.2  Gas Analysis
     Gas samples were analyzed by GC/FID for the identification and
quantification of hydrocarbons present.  The GC/FID's were equipped with
two 12'  by 1/8" stainless steel columns packed with 20% SP2100/.1%
Carbopack C.  Two columns Were necessary because Shimadzu GC/FID's use a
differential mode.  Samples were injected into the column by means of a
gas sample loop (Figure F-2).   If the sample required dilution a 500 cc
airtight syringe was used with a nitrogen dilutant.   Internal standards
were used to calculate the dilution ratio (Appendix B).
                                  7-3

-------
     Degradation studies on the gas samples were made on the first day's
testing.  Concentrations of components in the sample were measured over a
five and one half hour period.   During the first two hours,  degradation of
any constituent was less than 3% (Appendix B).   After two hours the error
from degradation became significant.   Therefore, all gas samples were
analyzed immediately after being sampled.
                                  7-4

-------
               8.0  QUALITY ASSURANCE AND QUALITY CONTROL

8.1  QUALITY CONTROL
     Quality control procedures were implemented for each type of analyses.
For gas samples, the Tedlar collection bags were filled with nitrogen
and analyzed by GC/FIO to ensure no residual compounds remained in the
bag.  The bags were also checked for leaks prior to each sample collection
by filling the bags to capacity and checking the seams and valves for
leaks.  For gas samples requiring dilution, a gas-tight 500 cc syringe
was used.  It was filled with nitrogen and analyzed by GC/FID prior to
each dilution to assure no residual hydrocarbons remained.  Each instru-
ment was calibrated daily with the following gas standards:  1000 ppm
hexane, 1300 ppm butadiene, 6.6 ppm benzene, 98.1 ppm benzene and 1060
ppm benzene.   A typical calibration curve for the benzene standards is
shown in Figure 8-1.  Calibration curves for the styrene in methylene
chloride standards and for the cyclohexane in methylene chloride standards
appear in Figures 8-2 and 8-3.  Instrument blanks were made after each
run by flushing the gas sampling system with nitrogen and confirming the
blank by GC/FID analysis.
     A degradation study was completed for one gaseous sample.
Concentrations of components in the sample were measured over a five-and-
one-half hour period (Table 8-1).   During the first two hours, degradation
of any constituent was less than 3% (Figures 8-4, 8-5, 8-6).   Therefore,
all gas samples were analyzed within the first two hours.  Degradation
rates of the cyclohexane audit samples were studied.   In a twenty-four
hour period,  degradation was negligible.
8.2  QUALITY ASSURANCE
     An performance audit was performed on the TRW procedures in the
field.  EPA provided three gas samples and three latex samples for the

-------
                    Table 8-1.   DEGRADATION STUDY SDO-A-3

Compound

Cyclohexane

Toluene

Ethyl benzene
Styrene


Retention
Time
2.42
2.73
3.79
4.96
5.8
7.15
7.93
9.15
11.01
14.65
12:17a
6414
15138
193533
596
1726
40010
9583
1
3745
690
2036
Area Counts
12:49
6177
14814
189519
783
1887
38895
9433
3720
525
2494
13:55
5923
14352
184543
540
1786
37737
9205
3324
571
2399
15.13
5674
13660
175205
500
1654
35563
8649
2977
531
2354
17:49
5413
13034
166829
487
1615
33098
8114
2992
525
2231
'Clock time.
                                 8-2

-------
      140
     120
      100
—    60
      40
      20
                                              •ean x
                                        ppm  area counts
            'slope
      S.D.
                                       8.2
                                       106
                                       495
  2.646
 36.639
156.957
0.0031 260.9
0.0029 3039
0.0031 4051
                   100        200        300

                       BENZENE CONCENTRATION  (ppm)
     400
   500
    Figure  8-1.  Benzene  Calibration  Curve  and Audit Results
                                  tt-3

-------
                                           G STYRENE
                                           Q AUDIT 131
                                           Q AUDIT 132
     1000
2000
3000
4000
SOOO
              Styrene Concentration (ppm)
Figure 8-2.  Styrene  Calibration Curve  and Audit Results
                            6-4

-------
    20





"ฃ IB

 K



 |  16





 ซr  14
 UJ
 ec



    12





    10





     8





     6





     4





     2
                           0 CYCLOHEXANE




                           (D AUDIT 111



                           177 AUDIT 112
                2000
4000
6000
8000
10000
                      Cyclohexane concentration (ppm)
       Figure  8-3.  Cyclohexane Calibration Curve and  Audit  Results



                                    8-5

-------
     190,000  -
     180,000  -
     170,000  -
QC
ft
     160,000  -
     150,000
                                       % LOSS
                                        13.8     O Styrene
 RT
9.15
                         Figure  8-4.   Hydrocarbon Degradation Study
                                      6-6

-------
                               NET LOSS
to

o
o
10,000

 9,000

 8,000


 7,000

6,000


5,000
   2,000
   1,000
                                                   Ethyl  Benzene
                                                   Butadiene
                                             O    Styrene
                                                   Unknown
                                                   Toluene
  RT
 7.93
 2.42
 9.15
14.65
 9.15
                   1        2        3         4        5         6
               Figure  8-5*   Hydrocarbon Degradation Study
                              8-7

-------




t-
o
ซz
ee

-------
audit, and supervised the test procedures.   The audit gases contained
propane, benzene, and hexane, and audit liquids of cyclohexane,  methyl
pentane, and styrene.  The samples were analyzed by GC/FIO and calibrated
as described in Appendix F.   The results of the audit samples are
presented in Table 8-2.
                                 8-9

-------
         Table 8-2.  RESULTS OF AUDIT SAMPLES FROM BORGER, TEXAS

Sample ID
BAL
XL
B
B
BAL
BAL
QAD
QAD
QAD
QAD
QAD
QAD
544
5
265
118
310
317
11
12
13
31
32
33
Measured
Concentration
79.1
2846
13.7
355
9.65
280
117883
\
3648a
2754
4612a
2087a
123493
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
Compound
Hexane
Hexane
Benzene
Benzene
Propane
Propane
Cyclohexane
Cyclohexane
As Cyclohexane
Styrene
Styrene
Styrene
aNumbers from laboratory analysis and are an average of analytical
 results.
 Concentrations expressed as a volume to volume basis.
                                 8-10

-------