United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 80-OCM-19
August 1980
Air
Benzene
Organic Chemical
Manufacturing
Ethylbenzene/Styrene
Emission Test Report
USS Chemicals
Houston, Texas
-------
SOURCE TEST AT UNITED STATES STEEL CHEMICALS
ETHYLBENZENE/STYRENE PLANT
HOUSTON, TEXAS
Contract No. 68-02-2812
Work Assignment 67
Project No. 80-OCM-19
Technical Manager: Winton Kelly
Prepared for:
U.S. Environmental Protection Agency
Emission Standards and Engineering Division
Emission Measurement Branch
Research Triangle Park, North Carolina 27711
TRW Incorporated
Environmental Engineering Division
Post Office Box 13000
Research Triangle Park, North Carolina 27709
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1-1
2.0 SUMMARY AND DISCUSSION OF RESULTS 2-1
3.0 LOCATION OF SAMPLING POINTS 3-1
4.0 SAMPLING AND ANALYSIS PROCEDURES 4-1
APPENDIX A SAMPLE CALCULATIONS A-l
APPENDIX B FIELD DATA SHEETS B-l
APPENDIX C FIELD GAS CHROMATOGRAPH STRIP CHARTS ... C-l
APPENDIX D FIELD LOG D-l
APPENDIX E LOW FLOW MEASUREMENT ASSEMBLY CALIBRATION . E-l
APPENDIX F AUDIT REPORT F-l
APPENDIX G PROJECT PARTICIPANTS G-l
-------
1. INTRODUCTION
During the week of June 23, 1980 personnel from TRW Environmental
Engineering Division, Energy and Environmental Analysis Incorporated
(EEA) and the U.S. Environmental Protection Agency (EPA) Emission
Measurement Branch (EMB) conducted tests at United States Steel Chemicals
(USSC) Ethylbenzene/ Styrene plant located in Houston, Texas.
This facility was tested in order to obtain and analyze samples to
provide data in support of possible National Emissions Standards for
Hazardous Pollutants (Benzene) and New Source Performance Standards
(Organic Chemical Manufacturing Industry).
Gas samples were taken of the combined gas inlet stream to the
superheater and the reboiler, the exhaust from the superheater, the
exhaust from the reboiler, the styrene finishing column vent to the
atmosphere, and the combined hotwell vent recovery system prior to the
natural gas ejector. Gas samples were analyzed for benzene, ethyl benzene,
toluene, xylene, styrene, C^-CQ alkanes, carbon dioxide (C02), oxygen (02),
nitrogen (N2), methane (CH4), and hydrogen (H2). The sampling locations
are identified by the symbols: CGI, SHO, RBO, SFCV, and OWSV, respectively.
The purpose of the testing at the combined gas inlet, the exhaust from
the superheaters and the exhaust from the reboiler was to determine the
destruction efficiency of benzene in combustion devices and benzene
atmospheric emissions levels. The styrene finishing column vent (SFCV)
was tested to determine the benzene and volatile organic carbon (VOC)
emissions to the atmosphere from an uncontrolled source of this type.
The purpose of testing the combined hotwell vent recovery system prior to
the natural gas ejector was to determine the composition of the recovered
gas stream.
1-1
-------
All sampling and analysis was performed at the USSC ethylbenzene/
styrene facility by TRW field personnel. Plant operating data was
obtained by personnel from EEA. The entire testing and analysis effort
was monitored and audited by the EPA/Emission Measurement Branch
Technical Manager.
1-2
-------
2. SUMMARY AND DISCUSSION OF RESULTS
2.1 GAS ANALYSIS RESULTS
The sampling and analysis data at USSC Houston were obtained from
five locations within the ethylbenzene/styrene facility. The first phase
involved the sampling and analysis of the gas stream composition at the
reboiler stack outlet, the superheater stack outlet, and the combined gas
inlet to the reboiler and superheater. The second phase involved sampling
and analysis at two column vent locations. These locations were the
styrene finishing column vent to the atmosphere and the combined hotwell
vent recovery system prior to the natural gas ejector.
During the first phase of testing, the three locations were sampled
simultaneously in order to determine the benzene mass removal efficiency
of the combustion processes in addition to the benzene emission
concentrations.
Summaries of the results of testing are presented in Tables 2-1
through 2-3. The average benzene emission results and removal efficiencies
are summarized in Table 2-4. The benzene emission concentrations from
the two heaters ranged from <0.1 to 0.34 ppmv. The average benzene
concentration in the fuel gas was 440 ppm. Using these results in a
combustion calculation (described in Appendix A) yields benzene removal
efficiencies ranging from 99.4 to 99.8 percent. The results are consistent
from run to run and there were no known sampling or process difficulties
during these tests.
The results of testing at the styrene column vent are presented in
Table 2-5. During sampling, condensate was observed in the sample line.
A sample of this liquid was analyzed and found to be essentially styrene.
No attempt was made to estimate the condensate volume or mass rate.
2-1
-------
Table 2-1. COMBINED GAS INLET GAS ANALYSIS RESULTS AT USSC-HOUSTON
ro
i
ro
RUN NO.
DATE
TIME
Species
analysis
C-la
C-2
C-3
C-4
C-4
C-5
C-5
C-6
C-6
C-6b
Benzene
Unknown
Unknown
Toluene
Unknown
Unknown
Ethyl benzene
Xylene
Styrene
CGI-1A
6/24/80
0900 MRS
R.T.
(min)
0.95
1.32
2.01
3.08
3.46
4.74
5.34
5.96
6.52
3.17
3.46
3.89
4.62
5.27
5.9
6.5
7.52
8.09
9.19
ppmv as
compound
751,944
26,247
7,719
865
1,531
428
417
115
90.3
53.1
425
55.4
78.3
153
31.2
<0.1
10.5
270
72.6
ppmv as
benzene
176,707
11,575
4,940
659
1,167
422
411
136
107
62.9
425
68.2
96.4
188
45.4
<0.1
15.3
393
83.0
CGI-1B
6/24/80
0920 MRS
ppmv as
compound
691,819
24,957
7,683
862'
1,546
9.22
2.83
438
440
52.3
421
53.2
73.9
150
30.1
<0.1
26.0
242
72.3
ppmv as
benzene
162,577
11,006
4,917
657
1,178
9.08
2.79
431
.433
62.0
421
65.5
91.0
184
43.7
<0.1
37.8
352
82.6
CGI-2
6/25/80
1000 HRS
ppmv as
compound
650,094
18,444
4,945
681
1,162
347
294
ND°
NO
42.9
425
48.5
77.8
169
34.9
13.4
29.3
258
86.6
ppmv as
benzene
152,772
8,134
3,165
519
885
342
289
NO
NO
50.9
425
59.7
95.7
208
50.8
19.5
42.6
375
99.0
CGI-3
6/25/80
AVERAGE
1400 HRS
ppmv as
compound
650,097
20,519
6,261
861
1,609
422
436
NO
NO
52.1
491
49.6
90.2
180
36.0
12.0
30.1
268
85.8
ppmv as
benzene
152,773
9,049
4,007
656
1,226
415
429
ND
ND
61.8
491
61.1
111
221
52.4
17.5
43.8
390
98.0
ppmv as
compound
685,988
22,542
6,642
817
1,462
409
397
138
133
50.1
440
51.7
80.1
163
33.0
6.35
24.0
260
79.3
ppmv as
benzene
161,207
9.941
4,257
623
1.114
403
391
142
135
59.4
440
63.6
98.5
200
48.1
9.25
34.9
378
90.7
Total hydrocarbons
by species
summation
790,505 197,501
728,878
182,550
677,152 167,532
681,500 170,103
719,509
179,422
Inertsc
H20, % by volume
N2, % by volume
02, % by volume
C02, % by volume
H2, % by volume
CH4, % by volume
TOTAL (%)
~0.0d
2.52
2.24
3.36
—
--
ND9
-0.0
0.94
0.4
3.31
19.8
74.5
98.95
-0.0
0.73
<0.1
2.89
17.9
75.4
Sc. no
U. ?(-
-0.0
0.57
<0.1
2.67
17.9
76.8
97.94
-0.0
0.75
0.13
2.96
18.5
75.6
97.94
Porapak Q column.
SP2100/0.1% Carbowax on 100/120 Supelcoport column.
GMolecular Sieve and Chromosorb 102 columns.
H20% not measured but assumed to be dry for reporting purposes.
-------
Table 2-2. REBOILER STACK OUTLET GAS ANALYSIS RESULTS AT USSC-HOUSTON
RUN NO
DATE
TIME
Species analysis
c-ia
C-2
C-3
C-4
C-5
C-6C
Benzene
i Toluene
CO
Ethyl benzene
Styrene
RBO-1 RBO-2
6/24/80 6/25/80
0960 HRS 1000 HRS
Retention" ppmv as ppmv as ppmv as ppmv as
time (rain) compound benzene compound benzene
0.95 . <0.1 <0.1 <0.1 <0. 1
1.33 <0.1 <0.1 <0.1 <0.1
1.97 <0.1 <0.1 <0.1 <0.1
3.41 <0.1 <0.1 <0.1 <0.1
6.49 <0.1 <0. 1 <0.1 <0.1
2.13 <0.1 <0.1 <0.1 <0.1
3.16 <0.1 <0.1 <0.1 <0.1
4.82 <0.1 <0.1 <0.1 <0.1
7.29 <0.1 <0.1 0.063 0.092
8.3 0.394 0.450 0.322 0.368
RBO-3
6/25/80 AVERAGE
1400 HRS
ppmv as ppmv as ppmv as
compound benzene compound
<0.1 <0.1 NDb
<0.1 <0.1 ND
<0.1 <0.1 ND
<0.1 <0.1 ND
<0.1 <0.1 ND
<0.1 <0.1 ND
0.345 0.345 0.115
0.324 0.399 0.108
0.470 0.683 0.178
0.526 0.601 0.414
ppmv as
benzene
ND
ND
ND
ND
ND
ND
0.115
0.133
0.258
0.473
Total hydrocarbons
by species
summation
0.394
0.450
0.385
0.460
1.67
2.03
Porapak Q column.
Not determined.
CSP2100/0.1* Carbowax on 100/120 Supelcoport column.
Molecular Sieve and Chromsorb 102.columns.
eH20 % by volume - saturated at analysis temperature 92°F.
0.815
0.979
Inerts
H20 % by volume6
C02 X by volume
CO % by volume
N2 % by volume
02 % by volume
H2 % by volume
CH4 % by volume
TOTAL (%)
~ 5.0
12.43
< 0.1
85.98
3.91
< 0.1
< 0.1
107.3
- 5.0
9.81
< 0.1
83.26
3.55
< 0.1
< 0.1
101.6
~ 5.0
9.81
< 0.1
83.10
3.17
< 0.1
< 0.1
101.1
~ 5.0
10.6
< 0.1
84.11
3.54
< 0.1
< 0.1
103.3
-------
Table 2-3. SUPERHEATER STACK OUTLET GAS ANALYSIS RESULTS AT USSC-HOUSTON
RUN NO. SHO-1
DATE 6/24/80
TIME •'*•'•' 0900 MRS
Retention ppmv as ppmv as
Species analysis time (min) compound benzene
C-la
C-2
C-3
C-4
C-5
C-6C
Benzene
Toluene
Unknown
Styrene
0.95 <0.1 <0.1
1.33 <0.1 <0.1
1.97 <0.1 <0.1
3.41 <0.1 <0.1
6.49 <0.1 <0.1
2.13 0.20 0.237
3.51 <0.1 <0.1
4.80 <0.1 <0.1
8.16 <0.1 <0.1
9.30 <0.1 <0.1
Total hydrocarbons
by species 0.20 0.237
summation
SHO-2 SHO-3
6/25/80 6/25/80
1000 MRS 1400 MRS
ppmv as ppmv as ppmv as ppmv as ppmv as
compound benzene compound benzene compound
<0.1 <0.1 <0.1 <0.1 ND
<0.1 <0.1 <0.1 <0.1 ND
<0.1 <0.1 <0.1 <0.1 ND
' <0.1 <0.1 <0.1 <0. 1 ND
<6.1 <0.1 <0.1 <0.1 ND
<0.1 <0.1 <0.1 <0.1 0.067
0.163 0.163 0.195 0.195 0.119
0.432 0.532 0.220 0.271 0.217
0.632 0.949 0.099 0.144 0.260
1.13 1.29 0.134 0.153 0.428
2.36 2.93 0.648 0.763 1.09
AVERAGE
ppmv as
benzene
ND
ND
ND
ND
ND
0.079
0.119
0.268
0.364
0.481
1.31
Inerts
H20, % by
N2, % by
02 , % by
C02, % by
TOTAL (%)
volume ~5.00e
volume 85.95
volume 5.96
volume 11.02
107.93
-5.00 -5.00
82.82 83.10
6.90 5.62
7.93 9.39
102.65 103.11
5.00
83.96
6.16
9.45
104.56
Porapak Q column.
Not determined.
CSP2100/0.1% Carbowax on 100/120 Supelcoport column.
Molecular Sieve and Chromosorb 102 columns.
CH20 % by volume - saturated at analysis temperature 92°F.
-------
Table 2-4. SUMMARY BENZENE REMOVAL EFFICIENCY
Run
Number
RBO-1
RBO-2
RBO-3
SHO-1
SHO-2
SHO-3
Moisture3
5.0
5.0
5.0
5.0
5.0
5.0
Oxygen
(% v/v dry)
3.91
3.55
3.17
5.96
6.90
5.62
Benzenec
(ppm wet)
<0. 1
<0. 1
0.345
<0.1
0.163
0.195
Benzene emissions Benzene inlet6
(ppmv @ 3% 02 dry) (ppmv)
<0.1 421
<0.1 425
0.366 491
<0.1 421
0.219 425
0.240 491
Dilution
Factor
8.73
8.41
8.36
9.79
10.33
9.61
Benzene
Removal
Efficiency^
99.8
99.8
99.4
99.8
99.6
99.6
Saturated at 92°-analytical temperature.
Determined by TCD analysis.
At stack conditions.
Equation A.3
eSee Table 2-1, Combined Gas Inlet (CGI).
See Appendix A - Table A-l.
9Equation A.4.
-------
Table 2-5. STYRENE FINISHING COLUMN VENT GAS ANALYSIS RESULTS AT USSC-HOUSTON
RUN NO.
DATE
TIME
Retention
Species analysis time (min)
C-la
C-2
C-3
C-4
C-4
C-5
C-5
C-6b
C-6
Benzene
Unknown
Unknown
Toluene
Unknown
Styrene
0.97
1.37
2.20
3.60
4.10
7.64
8.44
2.76
3.19
3.66
3.96
4.70
5.36
6.03
9.61
Total hydrocarbons
by species
summation
SFCV-1
6/26/80
ppmv as
compound
126,466
4,157
1,254
147
249
68.2
61.2
10.5
5.86
11.1
5.55
8.62
2.50
0.98
4,667
137,115
ppmv as
benzene
29,720
1,833
803
112
189
67.2
60.3
12.4
6.94
11.1
5.55
8.62
3.08
0.98
5,334
38,167
SFCV-2
6/26/80
ppmv as
compound
134,413
4,421
1,199
145
240
64.9
57.1
9.03
5.69
9.71
5.24
7.86
2.25
0.83
4,319
144,900
ppmv as
benzene
31,587
1,950
767
110
183
63.9
56.2
10.7
6.74
9.71
5.24
7.86
2.77
0.83
4,936
39,697
SFCV-3
6/26/80
ppmv as
compound
141,213
4,149
1,229
139
245
68.9
56.0
11.2
6.89
9.97
5.08
7.84
1.68
ppmv as
benzene
33,185
1,830
786
106
187
67.9
56.2
11.2
6.89
9.97
5.08
7.84
2.07
ppmv as
compound
134,031
4,242
1,227
144
245
67.3
58.1
10.24
6.15
10.3
5.29
8.11
2.14
AVERAGE
ppmv as
benzene
31,497
1,871
785
109
186
66.3
57.6
11.4
6.86
10.3
5.29
8.11
2.64
<0.1 <0.1 0.637 0.637
5,165
152,308
5,903
42,164
4,717
144,774
5,391
40.008
Inertsc
H20, % by
N2, % by
02, % by
C02, % by
H2 , % by
CH4, % by
TOTAL
volume
volume
volume
volume
vn 1 lime
volume
NDd
65.9
16.5
0.44
ND
15.1
97.94
ND
63.7
15.7
0.41
ND
15.2
95.01
ND
66.1
16.3
0.39
ND
15.0
97.79
ND
65.2
16.2
0.41
ND
15.1
96.91
Porapak Q column.
SP2100/0.1* Carbowax on 100/120 Supelcoport column.
cMolecular Sieve and Chromosorb 102 columns.
Not determined.
-------
During the testing at the styrene finishing column vent (SFCV)., the
volumetric flow rate was very low. The measurement performed indicated
that the flow was approximately 1 (one) standard cubic foot per
hour (scfh). The description and discussion of the volumetric flow
measurement is contained in Section 4.2 and Appendix E. The volume flow
rate results can only be considered estimates since it is suspected that
a leak was present in the measurement system and the results presented
could be low.
The results of testing at the combined hotwell recovery system are
presented in Table 2-6.
The sampling locations are described in Section 3, and the test
procedures are presented in Section 4.
2-7
-------
Table 2-6. COMBINED HOTWELL VENT RECOVERY GAS ANALYSIS RESULTS-AT USSC-HOUSTON
ro
oo
RUN NO.
DATE
TIME
Species analysis
C-la
C-2
C-3
C-4
Cm
-*t
C-5
C-5
C-5
C-5
C-6b
C-6
Benzene
Unknown
Toluene
Unknown
Unknown
Unknown
Xylene
Styrene
'*••*•.'
Retention
time (min)
0.93
1.34
2.12
3.53
3 "70
. Ij
4.10
6.30
7.60
8.40
2.70
3.10
3.40
4.60
5.25
5.93
6.40
7.47
8.00
9.15
OWSV-2
6/27/80
0800
ppmv as
compound
614.540
30,136
13,229
10,044
^n T
-------
3. LOCATION OF SAMPLING POINTS
Sampling locations are described in this section by the following
equipment groups:
• Reboiler and Superheater Systems
• Styrene Finishing Column Vent
• Combined Hotwell Recovery System
3.1 REBOILER AND SUPERHEATER SYSTEMS (CGI/RBO/SHO)
The USSC ethyl benzene styrene facility in Houston used a combined
gas stream to fuel both the direct-fired reboilers and the direct-fired
superheater. The combined gas inlet (CGI) stream was the first sampling
location. The combined gas stream was a composite of natural gas and
several process streams. The hotwell recovery system was one of the
process streams that merged to form the combined gas inlet stream. All
gas streams merged at the mixing drum prior to introduction at the
superheater and reboiler. The CGI sample was taken from a ground
location adjacent to the base of the superheater unit (Figure 3-1). A
representative portion of the gas stream was diverted through a bypass
around a control loop supplying the superheater. The combined gas line
was under positive pressure and consequently no evacuated can was necessary
for sample extraction. The gas flow to the sample train was regulated
by means of a flow meter attached to the sampling train assembly. The
line leading to the sampling train was thoroughly purged before
commencement of a sampling run.
This facility had two reboilers of similar design and capacity, A
and B. Reboiler A was selected for testing over Reboiler B because of a
more accessable sample port. Sampling at the reboiler stack outlet
(RBO) was conducted at approximately the 80 foot level. Access to the
3-1
-------
co
i
ro
From Combined Fuel Mix Drum
2" ID Pipe
SS Line
To Superheater
Control Gate
Valve
i;
P
Control Gate
Valve
4" ID Pipe
Flow Meter
Teflon Lines
Bag Can
Teiilar Bag
Figure 3-1. Combined gas inlet (CGI) sampling location.
-------
sampling location was by caged safety ladder (Figure 3-2). The reboiler
stack at the testing platform was approximately 8 feet in diameter. The
sample port was 2 inches in diameter and equipped with a gate valve and
V swagelok reducer assembly for adaptation to the sampling train. A
4 foot by h inch O.D. stainless steel probe was extended into the midpoint
of the flue gas stream. This sampling location was prepared by USS Chemical
personnel under a purchase agreement with TRW-Environmental Engineering
Division.
The superheater stack outlet (SHO) was the third sampling location.
This location was located at approximately the 60 foot level of the
superheater unit. The stack diameter at the testing location was 4 feet.
The sample location was prepared in a fashion similar to the reboiler
stack outlet location previously mentioned. Access to the location was
by caged safety ladder (Figure 3-3).
No difficulties were encountered in collecting the samples at these
locations once preliminary site preparation was completed.
3.2 STYRENE FINISHING COLUMN VENT (SFCV)
The styrene finishing column vent to the atmosphere was sampled to
assess the gas composition and flow rate from this uncontrolled source.
This testing location was atop the primary plant superstructure deck.
Access to the location was gained by stairwell (Figure 3-4). The gas
®
composition samples were taken from the vent by inserting the Teflon
sampling line approximately 3 feet into the vent duct and partially
sealing the vent to minimize dilution from ambient air. The flow rate
of the exhaust gas from this vent was extremely low. Initial attempts
to measure the flow rate from the SFCV with a vane anemometer proved
futile, because the flow rate of the source was below the threshold
necessary to propel the anemometer vanes. In order to approximate the
flow rate, a bag-manometer assembly was fabricated and is discussed in
Section 4.
3-3
-------
GO
I
Sample Port
80'
f-8'H
Reboiler A
Reboiler B
Figure 3-2. Reboiler stack outlet (RBO) sampling location.
-------
FIGURE 3-3. Superheater stack outlet (SHO) sampling location.
3-5
-------
co
i
01
7'
c
Heat Exchanger
Teflon Line
Evacuated
Can
12"
Figure 3-4. Styrene finishing column vent.
-------
3.3 COMBINED HOTWELL RECOVERY SYSTEM (OWSV)
The location of the sample point was one level below the top of the
main plant superstructure. Access was by way of a stairwell. The
sample was taken of the combined hotwell recovery system prior to the
natural gas ejector. The sampling location is diagramed in Figure 3-5
in order to identify primary system components of this unique recovery
system. Adaptation of the sample train to the vent gas line was made by
TRW personnel with the assistance of the plant (Figure 3-6). An airtight
and leakfree sampling system attachment to the gas stream was necessary
to prevent the formation of an explosive atmosphere within the plant
recovery system, due to the fact that the recovery system was under
10-15 inches of Hg vacumn. A single evacuated can at 28" Hg was
sufficient for the collection of a half-hour integrated bag sample. No
substantial sampling difficulties were encountered.
3-7
-------
co
oo
SECONDARY RECYCLE
RECOVERY COLUMN
STH
STYRENE
PRODUCT COLUMN
[REFLUX DRUM
HOT OPERATING
DURING TEST
RECYCLE OOHfil
[REFLUX DtUMJ
BENZENE/TOLUCNC
RECOVERY COLUMN
REFLUX DRUM|
Legend
STM - StoM
ATM - Atmosphere
jfor*- Condenser
OIL
WATER
Figure 3-5. Hotwell recovery system schematic.
-------
vo
0-60
PS1G
To Fuel Gas
Heater
fl
—
8° *
1 (X) 1
t ,
r '
n — fYr-
,
«
- 4;
T
^""^"
m^
— \
1" a. <.
1
II
111
Q" j
1
^^^^^^^^%
1
L m ^T>L ^~
*-/ — -VA) n
j
-1
—
T
1
14"
1
n
i <• _
M —
i
>r
^
1
ipr
X
I
.5".
Teflon
Sample Line
PS16
Natural Gas
Bag
Evacuated Can
fiLL
.2'
n
Fro*
Confined
Recovery
Condenser*
Figure 3-6. Hotwell recovery system (OWSV) sampling location and piping diagram.
-------
4. SAMPLING AND ANALYSIS PROCEDURES
4.1 SAMPLING SYSTEM
4.1.1 Evacuated Can Method
A modified Method 110 was chosen for use in collecting a sample.
The modification was the replacement of the vacuum pump with an evacuated
can. This system was chosen because of the explosion risk and safety
requirements of the plant.
The evacuated can method was used to collect a given quantity of
®
sample into a Tedlar bag. This method uses the negative pressure from
an evacuated can connected to a sample bag can as the mechanism for
obtaining a controllable sample flow. A diaphram pump is used to evacuate
the can, which is equipped with two self-sealing quick-disconnect valves,
to 29" Hg. A leak check is performed by connecting a vacuum gauge to one
of the quick-disconnect valves. If the pressure does not drop more than
1" Hg in 30 minutes, the can is considered to be leak-free. The equipment
is then transported to the sampling site and assembled according to the
Figure 4-1.
The sampling bags were checked for leaks before and after every
sampling run, by filling the bags with nitrogen (N2) and allowing the
®
bags to remain overnight. No leaks were observed. The Tedlar sample
bag is placed in the sample can and connected to the sample line that has
been purging at the site. The sample flow into the sample bag is obtained
by opening the valve between the 2 cans. The sample flow is monitored
with the flow meter. Adjustment of the valve will give the appropriate
sample flow.
The sample flow will remain constant until the evacuated can reaches
a low vacuum level. When the sample flow drops or the appropriate test
time is completed, the valve is closed between the cans and the sample
4-1
-------
PWK
KCOLE VM.VE
FLOWMETER
QUICK DISCONNECT
\
1
1
1
1
1
1
1
1
SAMPLING .
BAG
1
1
I
1
1
|
1
1
1
r ^
EVACUATED
CAN
y
Figure 4-1. Evacuated can sampling apparatus.
4-2
-------
bag disconnected from the sample line. The bag is capped off and removed
from the sample can. The bag is appropriately labeled and transported
immediately by the sampler to the laboratory for analysis.
4.1.2 Volumetric Flow Determination
During the test runs the gas flow data for the boiler and the
superheater were obtained by personnel from EEA. The gas flow rates
were recorded from control room instrumentation. The plant maintenance
personnel had calibrated the gas fuel flow meters to an estimated
10 percent accuracy prior to the sampling period. The data recorded is
presented in Table 4-1. The original data sheets as recorded in the
field are in Appendix B.
The volumetric flow of the exhaust gas to the atmosphere at the
styrene finishing column vent was very low and difficult to measure. A
Tedlar bag and manometer assembly was fabricated to approximate the
exhaust gas volumetric flow rate. Figure 4-2 is a generalized schematic
of the low flow measuring assembly. The complete assembly as utilized
in the field was reconstructed in the laboratory for post test calibration.
The calibration results are contained in Appendix E.
Data was collected during the sampling of the combined hotwell
recovery system to be used in an engineering calculation of the gas
volumetric flow. The data as collected by the field personnel are
presented in Table 4-2.
4.2 ANALYTICAL PROCEDURE
The test conducted at the United States Steel Chemical (USSC)
facility in Houston, Texas measured the volatile organic carbon (VOC)
emission levels of Cj-Cg alkanes, benzene, toluene, ethylbenzene, xylenes
and styrene. The samples were collected in hydrocarbon free Tedlar"'
bags and analyzed by gas chromatography/flame ionization detector
(GC/FID) for hydrocarbon concentrations and by gas chromatography/thermal
conductivity detector (GC/TCD) for hydrogen, methane, and fixed gas
concentrations. The instruments used were Shimadzu Mini 2 (equipped
with heated sampling loops), for hydrocarbon analysis a Shimadzu GC-3BT
for fixed gas analysis, and two Shimadzu C-RIA integrating recorders,
which were programmed to plot the sample peaks with their respective
4-3
-------
Table 4-1. CONTROL ROOM DATAC
Reboi
Fuel Flow
Date
6/24/80
Run #1
6/25/80
Run #2
6/25/80
Run #3
SCFH
118,000
118,000
118,000
118,000
118,000
120,000
120,000
120,000
122,000
122,000
120,000
120,000
120,000
120,000
120,000
SCMH
3,340
3,340
3,340
3,340
3,340
3,400
3,400
3,400
3,460
3,460
3,400
3,400
3,400
3,400
3,400
ler
Superheater
Stack outlet
temperature
°F
550
550
548
548
546
554
553
555
555
555
550
552
552
552
552
°C
288
288
287
287
286
290
289
291
291
291
288
289
289
289
289
Fuel
SCFH
32,800
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
32,000
flow
SCMH
930
910
910
910
910
910
910
910
910
910
910
910
910
910
910
Stack outlet
temperature
Op
456
460
460
458
456
458
462
462
462
462
462
462
462
464
463
°C
236
238
238
237
236
237
239
239
239
239
. 239
239
239
240
239
Data supplied by E.E.A. personnel.
4-4
-------
EZ3
4" Pipe
Reducer to 1/4"
Swagelock
Tedlar Bag
(evacuated)
Tygon Tubing (1/4")
Magnehellc
Ladder
Figure 4-2. Field low flow measurement assembly.
-------
Table 4-2. FLOW DATA COLLECTED AT COMBINED HOTWELL
RECOVERY SYSTEM3
Test run #
6/27/80
OWSV-1
OWSV-2
OWSV-3
OWSV-4
Ejector Inlet
pressure
(PSIG)
110
110
110
110
Ejector Outlet
pressure
(PSIG)
22.5
22.5
22.5
22.5
Recovered
Vapor line
vacuum
("Hg)
12.25
12.5
11.7
12.0
See raw data - Appendix B Field Data Sheets OWSV.
""See Figure 3-6 for location of gauges.
4-6
-------
retention times (R.T.), integrate the area under the peaks and calculate
the respective concentration in parts per million (ppm) based on
calibration constants. The calibration constants were determined from
daily calculations based upon commercially supplied CVC& and benzene
®
standards. The complete stripchart record from the C-RIA chromatapaks
are contained in Appendix C for reference. Each gas sample loop is
equipped with a vacuum system which incorporates an air flow meter,
adjustable between 0.1 and 0.5 liters per minute (1pm), so that a known
sample volume flows through the sample loop each chromatographic run. A
®
sample was drawn from the Tedlar sample bag with the vacuum system.
When ambient pressure conditions returned, the sample was injected linto
the column. After the chromatographic run, a three to five minute purge
of the sample loop and column with nitrogen provided a blank prior to
the next analysis. A record of blanks between analytical runs was
®
recorded by the chromatopak and may be inspected in Appendix C.
High purity gases were used to operate the instruments. Hydrogen
and air were mixed to provide the flame, helium was used as the carrier
gas, and nitrogen was used for purging purposes. Standard gases for
analysis include three benzene standards of different levels and a
cylinder containing C^C& alkanes of known ppm levels. (Methane 15.1
ppmv, ethane 14.6 ppmv, propane 15.6 ppmv, n-butane 15.2 ppmv, pentane
15.6 ppmv, hexane 15.9 ppmv.) The benzene standards had concentrations
of 6.6 ppm, 98.1 ppm, and 1,060 ppm in nitrogen. These standards were
run on the GC/FID prior to the field test to demonstrate linearity of
the instruments. Pure liquids of the other compounds of interest were
used to determine retention times. These compounds were toluene,
ethyl benzene, xylenes, and styrene. Disposable syringes were used to
sample the headspace above the liquid and inject the vapor into the gas
chromatograph during pre-field analysis preparation.
4.2.1 C,C6 Analysis
CiCe alkane analysis was conducted on a GC/FID operated in the
differential mode with a pair of Porapak Q columns, (61 x 1/8" O.D.
stainless steel). The column temperature was isothermal at 137°C (279°F)
with a carrier pressure 2 kg/cm2. At this temperature the retention
4-7
-------
time of C6 was approximately 18 minutes and low concentrations were not
measurable. The problem was alleviated by measuring the C6 alkane
concentration using the aromatic hydrocarbon procedure described in
Section 4.2.2.
4.2.2 Aromatic Analysis
The gas chromatograph (Mini-2) used for analysis of benzene, toluene,
ethylbenzene, xylenes, and styrene was operated in the differential mode
and equipped with two columns packed with 20 percent SP 2100/0.1 percent
on Carbowax 100/120 Supelcoport (61 x 1/8" O.D. stainless steel). The
column temperature was set at 75°C, isothermal. The helium carrier gas
pressure was set for 2 kg/cm2 (20 ml/min). The hydrogen and air flows
were set at the optimum level to achieve maximum detector response. To
minimize sample condensation the sample injection valve and sample loop
were contained in a valve oven and maintained at 150°C. This was necessary
because of high stack temperature and compound vapor pressure considerations.
This procedure alleviated any condensation problem and aided in reproduce-
ability in the analysis. Each sample was drawn through the sample loop
at 0.5 1pm for one minute. A three-to-five minute purge with nitrogen
was usually sufficient to bring the system to a blank level after each
run.
The inlet and process samples contain hydrocarbon levels which
overloaded the analytical column. To overcome this problem, each bag
sample containing elevated hydrocarbon concentrations was diluted with
nitrogen (N2) to approximately 50:1 through use of a large volume (500 cc)
gas tight syringe.
4.2.3 Fixed Gases
All the gas samples collected were analyzed for fixed gases using a
gas chromatograph equipped with a thermal conductivity detector. Two
6 foot x 1/8" O.D., stainless steel columns packed with Chromsorb 102
and Molecular Sieve were used in series to separate carbon dioxide (C02),
oxygen (02), nitrogen (N2), methane (CH4), and carbon monoxide (CO).
The analysis was recorded by the chromatopak . Stationary gas percentages
were calculated for each sample based upon daily determined calibration
factors from the commercially supplied stationary gas standards and the
4-8
-------
GC/TCO. During the analysis, finite levels of hydrogen in the gas
stream were anticipated. However, hydrogen (H2) was not detected, and
analytical modifications were necessary and are discussed below.
4.2.4 Analytical Problems
There were several minor analytical problems during the field
study. These problems include sample dilution, detector saturation and
hydrogen analysis with thermal conductivity. These problems were
anticipated from previous studies and were readily solved by adequate
preparation.
The high hydrocarbon levels encountered at the combined gas inlet
(CGI), the styrene finishing column vent (SFCV) and the combined hotwell
recovery system (OSWV) sampling locations required dilution. Previous
field studies showed good results using a large volume gas tight syringe.
The dilution ratio for each sample run was determined by comparing the
peak areas for C3-C6 compounds in the original and the dilute sample. A
ratio was also determined for benzene and the aromatic compounds in the
same fashion.
The analysis of both original and dilute samples at the CGI, SFCV,
and OWSV locations showed saturation for methane (Cx) and ethane (C:2)
compounds. The FID did not have a linear response to the high hydro-
carbon levels (>60 percent) encountered. Therefore, the apparent dilution
ratio (area counts of the original sample/area counts of the dilute
sample) was not used for methane and ethane. Instead, the average
dilution ratio of the C3-C6 compounds was utilized. In order to evaluate
the porportion of the sample that was methane (Cj) an alternative analysis
was used. High levels of methane was best measured by a thermal conduc-
tivity detector which is not subject to saturation at these high levels.
Therefore, the volume proportions of the primary gases in the samples
were determined by thermal conductivity as reported in the gas analysis,
summary (Tables 2-1 through 2-5).
During the initial analytical runs with the thermal conductivity
gas chromatograph, no hydrogen was evident on the chromatograms.
Subsequent inspection of the instrument setup revealed an apparent
masking of the hydrogen peak by other peaks (hydrocarbon) due to either
4-9
-------
improper column arrangement or decreased separation of components due to
column age. The apparent masking problem was remedied by subsequently
rearranging a new set of columns in an additional TCD instrument
(Shimadzu 6AM). A new single stainless steel column (6 feet by 1/8"
O.D.) packed with molecular sieve showed adequate resolution of the
hydrogen (H2) peak. The results of the hydrogen analysis for the plant
sample were comparable to results from TRW collected samples (integrated
®
Tedlar bag and grab stainless steel cyclinder) at the combined gas
inlet (CGI) sampling location.
4.2.5 Quality Assurance
The Environmental Protection Agency supplied three audit samples
for quality assurance purposes. They were treated the same as a regular
sample and analyzed on both GC/FID prior to sample analyses. The
chromatographic record of the audit samples analyzed prior to the field
samples; as well as the audit gases, that were reanalyzed during field
testing are contained in Appendix C. The audit report from the EPA/EMB
technical manager is contained in Appendix F.
4-10
-------
APPENDIX A
SAMPLE CALCULATIONS AND COMPLETE RESULTS
-------
Table A.I. COMBUSTION CALCULATIONS
Calculation Basis: 100 Moles Fuel
Let: a = volume % N2 in fuel
b = volume % C02 in fuel
c = volume % 02 in fuel
d = volume % H2 in fuel
e = volume % CH4 in fuel
L = volume % non-CH4 species in fuel
k = number of % non-CH4 species in fuel
m = hydrogen atoms in non-methane HC species
n = carbon atoms in non-methane HC species
1. Oxygen required for combustion, moles
k m.
02 = *sd + 2e + I L.(n. + ^ ) - c
2. Nitrogen with combustion air, moles
N2 = (79/21) 02
3. Carbon dioxide generated, moles
k .
C02=e+ I (L-n.)
i=l 1 n
4. Water generated, moles
k
H20 = d + 2e + I (L.m./2)
1=1 1
5. At stoichiometric air rates*, moles
02* = 02
N2* = a + N2
C02* = b + C02
H20* = H20
A-l
-------
Table A.I. Continued.
6. Since excess air is added and water is condensed prior to analysis:
Let: x = excess moles 02 added
y = moles H20 condensed
% N2e = N2 in exhaust sample
% 0e = 0 in exhaust samle
2 2
% 02e = 02 in exhaust sampl
% C02e = C02 in exhaust sample
7. Nitrogen balance:
* N e _ (N2* + 3.76 x)(100)
N2* + C02* + H20* + x + 3.76x-y
8. Oxygen balance:
e x (100)
" 2 N2* + C02* + H20* + x + 3.76x-y
9. Solving for x by eliminating y in the above equations:
% N2e - 3.76
* n e
* U2
10. Therefore y = N2* + C02* + H20* + 4.76x - -^-
%n
U2
11. At sample conditions, combustion products from 100 moles of fuel is
given by:
02S oxygen: x moles
N2S nitrogen: N2 + 1 + 3.76x moles
C02S carbon dioxide: C02 + b moles
H20S water vapor: H20 - y moles
TOTAL MOLES = SUM moles
12. Dilution ratio (DR) = |jj-j
C - C —
13. Mass removal efficiency = in out 100
A-2
Cin
-------
Table A.2. SAMPLE COMBUSTION CALCULATION
RUN # SHO-2 on 6/25/80
Assumption:
100 moles fuel basis
Inputs:
Inlet Fuel Composition
(moles)
N2 =
C02 =
H20 =
02 =
5H12 =
CeHi4 =
Benzene =
Toluene =
Ethyl benzene =
Styrene =
Xylene =
0.73
2.89
0
0
65.0094
1.8444
0.4945
0.1843
0.0641
0.00429
0.04735
0.02817
0.00427
0.00866
0.0258
H2 =100-1 inlet moles
H2 = 28.665
Outlet Gas from Analysis
(volume %)
N2e = 82.82
C02e = 7.93
H20e = 5.00
02e = 6.90
1. Oxygen required for combustion, moles
02 = %(28.665) + 2(65.0094) + 1.8444(3.5) + 0.4945(5) + 0.1843(6.5) +
0.0641(8) + 0.00429(9.5) + 0.04735(7.5) + 0.02817(9) +
0.00427(10.75) + 0.00866(10.5) + 0.0258(10.75) - 0
02 = 156.05 moles
2. Nitrogen with combustion air, moles
N2 = 79/21 (156.05)
N2 = 587.05 moles
A-3
-------
Table A.2. Continued
3. Carbon dioxide generated, moles
C02 = 65.0094 + 1.8444(2) + 0.4945(3) + 0.1843(4) + 0.0641(5) +
0.00429(6) + 0.04735(6) + 0.02817(7) + 0.00427(8) +
0.00866(8) + 0.0258(8)
C02 = 72.06 moles
4. Water generated, moles
H20 = 28.665 + 2(65.0094) + 0.5(1.8444)(6) + 0.5(0.4945)(8) +
0.5(0.1843)(10) + 0.5(0.0641)(12) + 0. 5(0.00429)(14) +
0.5(0.04735)(6) + 0. 5(0. 02817)(8) + 0. 5(0. 00427)(11) +
0.5(0.00866(10) + 0.5(0.0258)(11)
H20 = 168 moles
5. Stoichiometric Component Compositions Present
02* = 156.05 moles
N2* = 0.73 + 587.05 = 587.78 moles
C02* = 2.89 + 72.06 = 74.95 moles
H20* = 168 moles
6. x = excess moles 02 added
y = moles H20 condensed before analysis
= 71.3 moles excess 02
_ o 7C
3'76
"05
y = 587.78 + 74.95 + 168 + 4.76(71.3) - _ 71.3
y = 136.8 moles H20 condensed
11. At sample conditions
02S = 71.3 moles
N2S = 587.05 + 0.73 + 3.76(71.3) = 855.87 moles
C02S = 72.06 + 2.89 = 74.95 moles
H20S = 168 - 136.8 = 31.2 moles
SUM = 1033.32
A-4
-------
Table A.2. Concluded.
12. Dilution ratio = fgg = 10.33
13. Mass Removal Efficiency
100
Efficiency = 99.6%
A-5
-------
Table A.3. BENZENE REMOVAL EFFICIENCY
Benzene Concentration, ppmv @ 3% 02, dry
17.9
D-D Equation A. 3
Dd Bw (1-M) 20.9 - Od
where:
B. = Benzene concentration ppmv @ 3% 02, dry
B = Benzene, ppmv, wet
W
20.9 = Oxygen (% v/v) in ambient air
0, = Oxygen, % v/v, dry
M = Moisture fraction, —
100
Equation A. 4
MR(%) = CIN " COUT
-------
Table A-4. BENZENE MASS REMOVAL EFFICIENCY
Removal Efficiency
Run # Calculations (%)
^ 0. 0421 -l(10-»)(9. 791 X100= ^
SH02 0.0425-1.6jfllO-*)(10.33) X100 = "'6
SH03 - - X100= 99'6
RB01 -- X100= 99'8
RB02 - " Xl°°= 99'8
RB03 - - X100= 99'4
A-7
------- |