&EFA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 79-CKO-16
September 1979
Air
Iron and Steel
(Coke Oven Battery
Stack)
Emission Test Report
C F & I Steel Corporation
Pueblo, Colorado
-------
Emission Testing at a
By-Product Coke Plant
(Battery D Stack)
C.F. & I. Steel Corporation
Pueblo, Colorado
Prepared for the
U.S. Environmental Protection Agency
Emission Measurement Branch
Research Triangle Park, North Carolina 27711
Prepared by
Clayton Environmental Consultants, Inc.
25711 Southfield Road
Southfield, Michigan 48075
EMB REPORT NO. 79-CKO-16
Work Assignment 16
Contract No. 68-02-2817
-------
FOREWARD
Two firms prepared this report under contract'
to the U.S. Environmental Protection Agency; therefore,
it is presented in two sections.
Section I was prepared by Clayton Environmental
.-
Consultants, Inc., Southfield, Michigan, and includes
test results for particulate, sulfate, and chloroform/
ether extractables; NOX (Method 7); continuous monitor-
ing of CO, 02, and N0x; 02, CO, C02 (Method 3); and,
visible emission data for the Battery D stack exhaust.
Section II was prepared by TRW Energy Systems
Group, Redondo Beach, California, and contains
the benzo(a)pyrene (B(a)P) sampling data, and summary,
which immediately follows Appendix H of Section I.
-------
TABLE OF CONTENTS
Page
SECTION I
List 'of Tables i
List of Figures ii
•*•
1.0 Introduction ' 1
2.0 Summary and Discussion of Results 3
3.0 Process Description and Operation 22
4.0 Location of Sampling Points 23
5.0 Sampling and Analytical Procedures 27
APPENDICES .
A. Project Participants
B. Field Data Sheets
B-l. Particulate Test Data Sheets
B-2. Sampling Summary Data
B-3. Visible Emissions Data Sheets
B-4. Summary of Visible Emissions
B-5. Nitrogen Oxides Data Sheet
B-6. Process Operating Logs and
Charts
C. Continuous Monitoring Data
C-1. Carbon Monoxide
C-2. Nitrogen Oxides
C-3. Oxygen
D. Example Calculations
E. Opacity and Carbon Monoxide Correlation
Data
-------
TABLE OF CONTENTS (CONTINUED)
F. Detailed Summary of Sampling and
Analytical Procedures
F-l. Non Dispersive In.frared
(Continuous Carbon Monoxide)
F-2. Continuous Nitrogen Oxides
*-
F-3. Continuous Oxygen
G. Calibration Data
H. Weights by Fraction
SECTION II
-------
SECTION I
-------
LIST OF TABLES
Table Page
2.1 Particulate Concentrations and 5
Emi s s ion Ra tes
2.2 Sulfate Concentrations and Emission 7
Rates
2.3 Sulfate as a Percent of Particulate, 8
by Weight
2.4 Nitrogen Oxides Concentrations and 10
Emission Rates
2.5 Summary of Correlation Results 18
206 Exhaust Gas Composition Analysis 20
-------
LIST OF FIGURES
Figure Page
2.1 Graphic display of continuously H
monitored data - Run 1
2.2 Graphic display of continuously 12
monitored data - Run 2
2.3 Graphic display of continuously 13
monitored data - Run 3
4.1 Sampling location on waste heat 24
stack
4.2 Location of sampling points 25
5.1 Particulate sampling train 29
5.2 Orsat gas sampling train 36
5.3 Sampling train for continuous 39
monitoring of carbon monoxide
5.4 Combination sampling trains for 42
continuous monitoring of nitrogen
oxides and oxygen
5.5 Sampling train for nitrogen oxides 44
grab samples
11
-------
1.0 INTRODUCTION
The U.S. Environmental Protection Agency retained
Clayton Environmental Consultants, Inc., to determine
particulate and various gaseous emissions from the
battery stack of Coke Oven Battery D at C.F. & I.
*•
Steel Corporation in Pueblo, Colorado. The results
of this study will be used in research and development
efforts for supporting New Source Performance Standards
for coke oven battery stacks in the iron and steel
industry. This study was commissioned as EMB Project
No. 79-CKO-16, Contract No. 68-02-2817, Work Assign-
ment 16 .
The testing program included the following:
(1) Triplicate samples to be analyzed for
particulate, sulfate and chloroform/ether
extractables (of the impinger solution);
(2) Integrated bag samples for Orsat analyses;
(3) Continuous monitoring of carbon monoxide,
nitrogen oxides and oxygen during the
particulate runs;
(4) Grab samples for nitrogen oxides;
(5) Visible emission observations
for the duration of each particulate
sample run.
Auxiliary data included exhaust gas temperatures
and flowrates, as determined from the traverses.
-------
The study was conducted on August 7, 8, and 9,
1979 by Clayton Environmental Consultants, Inc. with
the continuous nitrogen oxides and oxygen monitoring
efforts subcontracted to York Research Corporation
of Denver, Colorado. Project participants are
listed in Appendix A.'
- 2 -
-------
2.0 SUMMARY AND DISCUSSION OF RESULTS
Results of the emission study are presented in
Tables 2.1 through 2.5. Tables 2.1 and 2.2 present
the filterable and total concentrations and emission
rates of particulate and sulfate, respectively.
The filterable fraction (front half) includes the
particulate from the probe, front portion of the
filter holder and the filter. Total particulate
includes the filterable fraction plus the particulate
in the rear half of the filter holder, the impingers,
and connecting glassware up to the silica gel impinger.
Concentrations are expressed as grains per dry standard
cubic foot (gr/dscf) and milligrams per dry standard
cubic meter (mg/dscm). Emission rates are expressed
as pounds per hour (Ib/hr) and kilograms per hour
(kg/hr).
Table 2.3 presents sulfate expressed as a percent
of particulate by weight. Sulfates are expressed as
sulfuric acid (including sulfur trioxide). Nitrogen
oxides concentrations and emission rates are presented
in Table 2.4. Concentrations are expressed as parts of
nitrogen dioxide per million parts of air (ppm) and
the corresponding emission rates as Ib/hr and kg/hr.
Table 2.5 presents the results of the exhaust gas
composition (Orsat) analyses.
- 3 -
-------
All field data sheets and continuous monitoring
data are included in Appendices B and C, respectively.
Appendix D presents an example of calculations used
to interpret the data collected.
PARTICULATE RESULTS
The measured filterable concentrations of partic-
ulate from Battery D, shown in Table 2.1, ranged from
0.035 to 0.045 gr/dscf (80.4 to 104 mg/dscm) and
averaged 0.039 gr/dscf (89.9 mg/dscm). Concentrations
o>f total particulate ranged from 0.043 to 0.064 gr/dscf
(98.7 to 147 mg/dscm), and averaged 0.051 gr/dscf (118
mg/dscm).
Emission rates of filterable particulate ranged
from 6.76 to 7.21 Ib/hr (3.07 to 3.27 kg/hr) and
averaged 6.93 Ib/hr (3.15 kg/hr). Total particulate
emission rates ranged from 8.38 to 9.54 Ib/hr (3.80
to 4.33 kg/hr), and averaged 9.02 Ib/hr (4.09 kg/hr).
Generally, the data reflect good reproducibility
(within 10 and 25-percent of the mean, based on concen-
tration) .
Run 1 results show a higher concentration of
filterable and total particulate than Runs 2 and 3.
The emission rates of filterable and total particulate,
however, show lower filterable emissions and higher
total particulate emissions than Runs 2 and 3. The
flowrate during Run 1 was approximately 23-percent
lower than during Runs 2 and 3, which contributes to
- 4 -
-------
TABLE 2.1. PARTICULATE CONCENTRATIONS AND EMISSION RATES
Sample
Number
Date
1979
Stack Gas
- Parame ter s
Fl owra te
d s e f m .
Temp
F
Concentration
Filterable
gr/d sc f
mg/d scm
Total
gr/dscf mg/dscm
Emission Rate
Filterable
Ib/hr
kg/hr
Total
Ib/hr
kg/hr
1 8
2 8
3 8
-7
-8
-9
17,400
22,600
22,700
440
430
420
0.045
0.037
0.035
104
85.2
80.4
0.064
0.047
0.043
147
108
98.7
6.76
7.21
6.83
3.07
3 -.27
3.10
9.54
9.13
8.38
4.33
4.14-
3.80
Average
20,900
430
0.039
89.9
0.051
118
6.93
3.15
9i02
4.09
-------
the differences in the calculated emission rates
between these runs.
SULFATE RESULTS
Table 2.2 presents the measured filterable concen-
trations of sulfate, which was 0.012 gr/dscf for all
.»-
three runs (ranging from 26.5 to 28.5 mg/dscm and
averaging 27.2 mg/dscm). Concentrations of total
sulfate ranged from 0.021 to 0.050 gr/dscf (47.6
to 115 mg/dscm), and averaged 0.032 gr/dscf (72.3
mg/d scm).
Emission rates of filterable sulfate ranged from
1.85 to 2.26 Ib/hr (0.841 to 1.03 kg/hr), and averaged
2.12 Ib/hr (0.964 kg/hr). Total sulfate emission
rates ranged from 4.02 to 7.49 Ib/hr (1.83 to 3.40
kg/hr), and averaged 5.37 Ib/hr (2.44 kg/hr).
S-ulfate, as a percent of the filterable particulate
by weight (Table 2.3), ranged from 27.4 to 33.1-percent
and averaged 30.5-percent. The total fractions ranged
from 44..1 to 78.5-percent and averaged 59.2-percent.
Filterable sulfate concentrations are very
reproducible over the three sample runs (0-percent
variation about the mean). The emission rates of
filterable sulfate are higher in Runs 2 and 3 than
in Run 1 despite the similarity in sulfate concentra-
tions, due to the higher flowrates which were measured
in these runs. Total sulfate concentrations also
- 6 -
-------
TABLE 2.2. SULFATE CONCENTRATIONS AND EMISSION RATES
Sample
Number
Date
1979
.-Stack Gas
Parame ter s
Flowrate Temp
d sc fm F
Concentration
Filterable
gr/dscf mg/dscm
Total
gr/dscf
mg/d scm
Emission Rate
Filterable
Ib/hr
kg/hr
Total
Ib/hr
kg/hr
1 8
2 8
3 8
j
-7
-8
-9
17,
22,
22,
400
600
700
440
430
420
0.012
0.012
0.012
28.5
26.5
26.6
0
0
0
.050
.021
.024
115
47.6
54.3
1 . 85
2.24
2.26
0.84T
1.02
1.03
7.49
4.02
4.61
3.40
1,83
2.09-
Average
20,900
430
0.012
27.2
0.032
72.3
0.964 5..37
2.44
-------
TABLE 2.3. SULFATE AS A PERCENT OF PARTICULATE
(BY WEIGHT)
00
Sample
Number
1
2
3
AVERAGE
Filterable
27.4
31.1
33.1
30.5
Total
78.5
44.1
55.1
59.2
-------
show reproducibility for Runs 2 and 3 (with a 12-percent
variation), but Run 1 shows twice as much total sulfate
as Runs 2 and 3. A similar relationship exists in terms
of total sulfate emission rates.
Total sulfate, as a percent of total particulate
by weight, is 58-percent higher in Run 1 than the
average of Runs 2 and 3. This fact, coupled with the
higher total particulate concentrations and emission
rates encountered in Run 1 versus Runs 2 and 3, and the
differences in measured flowrates in these sets of
data, suggests a possible difference in the process
operating conditions between Run 1 and Runs 2 and 3.
NITROGEN OXIDES RESULTS
Nitrogen oxides concentrations for the Method
7 flask grab samples (Table 2.4) ranged from less
than 7.66 to 81.9 ppm, and averaged 43.9 ppm.
Emission rates ranged from less than 0.954 to 13.3
Ib/hr (less than 0.433 to 6.04 kg/hr), and averaged
7.06 Ib/hr (3.21 kg/hr). These averages do not
include the "less than" values.
These grab sample results do not closely agree
with the continuously monitored measurements of nitrogen
oxides which are shown in Figures 2.1, 2.2, and 2.3.
These measurements all fall within a range between
80 ppm, which occurred during Run 3 and 130 ppm, which
occurred during Run 2. Grab Sample 2 does fall within
this range, but Samples 1, 3, and 4 are all much lower
- 9 -
-------
TABLE 2.4. NITROGEN OXIDES .CONCENTRATIONS AND EMISSION RATES3
Sample
Number
Date
1979
Concentrations
ppm
Emission Rate
Ib/hr
kg/hr
1 8
2 8
3 8
4 8
-7
-8
-8
-9
Average
< 7
81
35
14
43
.66
.9
.8
.0
.9
<0
13
5
2
7
.,954
.3
,59
.28
.06
<0.
6.
2.
1.
3.
433
04
54
04
i
21
Method 7 flask grab samples (as nitrogen dioxide).
'Average based on Samples 2, 3, and 4. •••
- 1-0 -
-------
./Figure 2.1. ^Grapjiic 'display of continuously monitored data.
Ui-u jfrniar-ttjitrl!-! U±cttt
Stack gas temperature, F
|lfi'[i||'oxygen by volume, ppm
' ,;j; : ; !;•; i;:t? -UU;/:;U::j4:;rUr-
* - . 1 ..,*)* .--J j [• j+-tT-i f - . f f i-;4~H |Xu
- !-:t.: :j ii:TH±Hjiirr^H-^-H—*+.
£ 2 i [•! [' v:•• Jil'USiHlliilSiL
srr-Ea ~-i } TCyjpf .^-m ^j^-^r^a«i- vs -——re—; ;—-
gfEife,^ |[pr FJTj
ifMtl lilil,/11.H' ! I frf JJilj
i--l4.lt4.ll -it U / . it ' --L 1J»U U U- -IJJI !-L4-!- JUvl i t-1
-S itfrrt, 11 .,,/_ -jj^... | I. ..| 171 i l-i. .U-l ii 4-M4-t- iJJ-tlj-i-l
sffittffiti^K^ w
jiiitiliitittb. liii i '"^-r!rRtf4 -H HTtj+l-rH,- • R-ri
l±rit^Iyjjiffli-i-i-i-l-^U ."; [j'Hi'mTJl-m+H't-i-' • ILJ.n'
S^ppr^w^w
.rJ-H-f-jf rrj~-^, I ' j | t-t_-_j
witFNitrogen oxides by volume, ppm
. j. i. . l j i. , „ _i L _ ^,_
4 -, : : : !.J. i-1 J ' ' ' I J 4 11 I l J ! l ili l-.M^i-4-l..
I^Opacity^ percent
'.!( 1"" ;***- •-.--t-^t-t-—•-•-(-*-'• »-i i * f r-t--r-pTi n 1777T[T!T'"L"~Jr ll-P-l;
|^|4|_(No charging a c t i v i t y ) ffp3|||:R|
-f 3« i. Mfi n^U:,i;.-.:i-:.ria-'.Nviitii^TlJ^PSi
feritjrifirfrriiui-.j tiii iitrti^TOJt^lf'fWm^:
.j ] . .|—i-^-.-i-)-.i j.; ,H- i-i-T-i-) -o.-^ t4-»-i i I I i < I I'* M-;-f-t-
S^xHCO concentration,
-•i-H-£ -r'-f J 4 ^
1710
740.
1800
1830 . 1900
Run_ JL^ Samp 1 i n g_ time
19TO'
2000
2020
• /11
-------
Figure 2.2. :Graphic display _of -continuous ly monitored data,
+tps_tack_ gas tempera
by volume, ppin rjtj
-ttttn-1-H itu4 i±k :U -H !Uttt^p-^
-------
Figure 2.3. Graphic^display of 'continuously monitored data.. .
u>
400
30Q
200
12.5
7-.5
1.10
" \
. 100
'. 90
Port change) £
Stack gas temperature, F
]:.:-::p::!:r)"i:n::^:j;-'
-'••fy^mi^
L^it tt-tU'il :llitt:r±riUirCx^vl
/£
^jlr..^) jj. *.! ' .-.l-.X-.._j---.- i _.* ' r^t -* M '._"111-:*: i.rj"'
1!-j :!,i :!: I'fHiH t
uiillil I •:-tir'fFM-''-,
; -11 •
[Mir
]]^^
Charging :
1120
1140
1200
~12 3 O// 1450.
1510
1530
1600.^
-Run—3._Samp.Li.ng .-time
-------
than the continuous monitor recordings. The number
of grab samples taken was not sufficient to determine
if a true difference or apparent trend existed between
the continuous monitor and the grab sample results.
CONTINUOUS NITROGEN OXIDES MONITORING
Nitrogen oxides concentrations, as measured by
the continuous chemiluminescence method, ranged from a
low of 80 ppm during.Run 3 to a high of 130 ppm during
Run 2 (Figures 2.1, 2.2, and 2.3). These concentra-
tions appear to fluctuate independently of temperature,
oxygen, carbon monoxide, and opacity.
CONTINUOUS OXYGEN MONITORING
The oxygen concentrations recorded by the continuous
monitoring method ranged from a low of 8.5-percent during
Run 3 to a high of 12.0-percent during Runs 1 and 2.
These data correspond favorably to the oxygen levels
measured by the Orsat method.
..Oxygen levels remained relatively stable and
appear to fluctuate independently of carbon monoxide,
nitrogen oxides, opacity, and stack temperature fluctua-
tions. Figures 2.1, 2.2, and 2.3 also display the
recorded fluctuations of the oxygen levels.
CONTINUOUS CARBON MONOXIDE MONITORING
With the exception of spiking, which coincided
with the regularly-occurring coke oven gas reversals,
carbon monoxide measurements were generally below 300
-------
ppm, and usually remained near zero. Carbon monoxide
concentrations are graphically displayed in Figures
2.1, 2.2, and 2.3 for Runs 1, 2, and 3, respectively.
The two-way radios used by the C.F. & I. coke
oven battery workers caused occasional "radio
interference" problems with the NDIR analyzer,
which limited the precision attainable from the
strip chart recordings, especially in the concentra-
tion range encountered.
There seems to be no relative correlation
between carbon monoxide and1 nitrogen oxides, oxygen,
and temperature data. Stack gas opacity, however,
seems to correspond with and closely track carbon
monoxide fluctuations, thus a linear regression and
correlation analyses were performed on the data.
Statistical Analysis of Data
The time-concentration curves (Figures 2.1, 2.2,
and 2.3) were reduced in the following manner: The
carbon monoxide strip chart continuous readings were
reduced to 15-second point readings to correspond
with discrete 15-second opacity readings. The CO
readings were rounded to the nearest,100 ppm and
those less than 100 ppm were interpreted to be less
than the limit of detection. A data file was then created
using corresponding opacity and CO readings at a given
point in time. One such data file was created for
each sample run.
- 1-5 -
-------
Since subjective observations indicated that CO
concentration peaks were generally preceded within a
few minutes by a rise in opacity, a computer program
was devised which would accommodate and adjust the
data set pairings for any given lag time. The lag
**
time indicates the time, in minutes,, before and after
(negative or positive) an opacity reading to the
associated CO reading. Each data file was then run
through a linear regression program and correlation
routine to determine if a significant relationship
existed betv?een the data for a given lag time.
Different lag times were used to determine the optimum
(maximum) correlation coefficient (r), beginning
at whole minute intervals then reducing to quarter
of a minute intervals. This usually required five
to ten runs per sample.
Several problems necessitated altering the data
inputs to accommodate a more realistic analysis. For
example, steam exiting the quench tower occasionally
obscured the battery stack emissions. Therefore, for
these points in time, there would be CO readings but
no associated opacity readings. Thus, these data
could not be counted as a valid data set. The number
of complete pairs of data available for correlation
then, i.e., the number of. data sets used, was less
than the total number of pairs first described
(above). Therefore, a new data file was created
- -16 -
-------
based on the optimum time lag (defined by the maximum
"r") and only complete data sets were utilized in
the statistical analysis.
Re sults
Table 2.5 presents the results of the correlation
*•
analyses. A different time lag produced the optimum
correlation coefficient (r) for each run (variation:
-3.5 to 3 minutes). The results for each sample
yielded a r-value of about 0.7. .During the second
half of Run 2, the CO readings were suspect due to
very low readings, thus a second program iteration
was made using only the first half of the test.
This resulted in the much higher coefficient of 0.86.
Appendix E presents the distribution of data pairs
for each run.
Although continuous nitrogen oxide and oxygen
data was available, the resolution of overall magnitude
of oxygen concentrations was not sufficient to resolve
the data over as many points as was needed. From the
graphic summaries, however, no apparent trends seemed
to exist. Therefore, this was the extent of the
statistical analyses performed on this data.
Comparison with Previously Collected Data
The data from the C . F .". & I. study was quite different
than that previously collected at another battery stack
location. Concentrations of carbon monoxide measured during
- 17 -
-------
TABLE 2.5. SUMMARY OF CORRELATION RESULTS
oo
Sample
Number
1
2
3
No.
of
Data
Sets
160
271
173b
81
% of
Data
Used
100
100
100'
100
Corre lation
Coefficient
(r)
0.6802
0.6958
0.8638
0.7403
Linear Regression Lag Time,
Equation Minutes3
CO-= 29.6 op + 69.8 -3.5
CO = 57.2 op +121
2
CO = 42.7 op + 114
CO = 24.4 op + 201 3
Minutes from opacity reading to the carbon monoxide reading.
Run using only the first half of the sample.'
-------
the earlier study ranged from 100 to 1500 ppm,
while at C.F. & I. the greatest recorded concen-
tration was 1100 ppm. Similarly, the maximum
opacity at the other facility was 50-percent
while at C.F. & I. the maximum was 20-percent,
•»*
with most of the readings at 0-percent. As may
be seen from the distribution of data pairs (Appendix
E) for all runs, the majority of the readings at
C.F. & I. were .less than 150 ppm and either 0 or 5-
percent opacity. This did not offer the wide varia-
tion in paired data sets as was present in the earlier
study, thus outlying data sets did not suggest the
need for elimination. Since no charge times were
available as additional input, this relationship
could not be explored for C.F. & I.
EXHAUST GAS COMPOSITION
Table 2.6 displays the results of the exhaust gas
composition analysis using Method 3. Determinations
of carbon dioxide, oxygen, and carbon monoxide content
were made for each of the three sample runs. Moisture
content is also presented and shows an average of
13.9-percent.
VISIBLE EMISSIONS
Visible emissions from the Battery D stack were
recorded for the duration of each particulate sample
run, with two exceptions. Readings were terminated
- 19 -
-------
TABLE 2..6. EXHAUST GAS COMPOSITION ANALYSIS
Sample
Number
Moisture
Content
percent
Exhaust Gas Composition, Dry Basis
percent
Carbon
Dioxide
Oxygen
Carbon
Monoxide
Nitrogen
and Inerts
j
1
N>
b
i
1
2
3
13.
13.
14.
8
7
2
4.
5.
4.
2
2
9
9.
9.
10.
1
5
7
<0ol
<0. 1
<0. 1
860
85.
84.
7
3
4
Average
13.9
4.8
9.8
85.5
-------
within minutes of the completion of Run 2 due to
insufficient sunlight. Overcast conditions permitted
about 1-hour of observation during Run 3.
The opacities recorded during these runs were
consistently low, with the exception of occasional
•*-
peaks and regularly occurring peaks which coincided
with rises in carbon monoxide concentrations and with
the coke oven gas reversals in Battery D. These
reversals occurred every half-hour at approximately
20 and 50 minutes past the hour. The peaks in opacity
occurred approximately four minutes following carbon
monoxide peaks during Run 1 and three minutes prior
to carbon monoxide peaks in Runs 2 and 3. Figures 2.1,
2.2, and 2.3 graphically depict the fluctuations in
opacity during each run.
- 21 -
-------
3.0 PROCESS DESCRIPTION AND OPERATION
To be supplied by EPA.
- 22 -
-------
4.0 LOCATION OF SAMPLING POINTS
At the sampling location elevation, access to the
10.2-foot I.D. waste heat battery stack was obtained
through two of three sampling ports positioned approxi-
mately five stack diameters downstream from the under-
ground, primary underfire flue duct and about 25 diameters
upstream from the stack outlet. Two of the three
ports had been previously installed through the four
foot thick stack wall, at a 180-degree angle about
the stack circumference, while the third port was in-
stalled perpendicular to these ports specifically for
this testing program. Figure 4.1. characterizes the
sampling location relative to the Coke Oven Battery D
and stack. A total traverse of thirty-six sampling
points, equally divided between two sampling ports
spaced at 90 degrees, was employed during the testing
program. Figure 4.2 depicts the location of each sam-
pling point with respect to the inner stack wall.
The number of sampling and traverse points chosen
afforded suitable velocity traverse data, considering
the very low (natural draft) flowrates. The four
outermost traverse points' sampling times were incor-
porated into the next inwardmost points due to their
- 23 -
-------
Figu
re 4.1
Sampling location on wasl
Coke
Oven
Ba ttery
D
'
te heat
Are
of
c r o
sect!
FT 211 T P
j_ £ u. J. c:
Door
car
Tracks
n n
stack
/
f l
/^
a
s s
on ' A-A1
4.2.
Quenclr
car
t racks
25 ft.
n n
\
\
•si
^1
<
I
!
^
/
'
1
I
18 ft. *
i 1 ' H
I I i!
i ! ! i
i 1 !
i ! i ':
1-— '
,
i
1
1
'
II l! II .
- 't -U ll —
i !
i
! •'
!
i
'
.
-4 1—
/
f
T_
_l
*""\
-^
1
I
!5
f
(L
C
-»-
3
\
0'
A
/rr\x
®\
ports
\
J Samplir
_ _] P 1 a t f o i
0'
Ground
' Elevation
\
Primary waste heat gases
Not to scale
-------
!
ro
Ln
1
Po int
. Numbe r
1
2
Distance from wall
inches
1.7
5.4
3 j 9.2
4 ! 13.3
5 1 17.9
6
7
8
9
10
11
12
13
14
23.0
cm
4.4
13.7
23.3
33.9
45 .5
58.4
28.9 1 73.4
36.2
46.8
75.6
86.2
92.0
118.8
192.1
218.9
93.5 ! 237.5
99.4
252.5
104.5 I 265.4
15 j 109.1
16 ! 113.2
17
117.0
18 j 120.7
277.0
287.6
297.2
306.5
i i
Port
10 . 2-'
Catwalk
Grating
Figure 4.2. Location of sampling points
Section 'A-A
Ladder
-------
being positioned too close to the stack port liners
to obtain representative velocity pressures. These
port liners were not flush with the stack inner
wall.
Throughout the testing program, difficulty was
encountered when changing ports and with the support
of the 16-ft probe. When the probe was largely
outside of the stack, .its weight placed consid^-
erable pressure on the front part of the probe,
forcing the total length of the probe to bow. This
bowing eventually caused the nozzle to catch the
inside of the stack wall, making the probe intractable
unless first disassembled. This problem precipitated
the modification of the sampling nozzle after comple-
tion of the first test. With the approval of the
EPA Technical Manager, a portion of the "elbow" type
nozzle tip was removed and re-tapered. This allowed
removal of the probe from the port without disassembly,
- .26 -
-------
5.0 SAMPLING AND ANALYTICAL PROCEDURES
A detailed summary of sampling and analytical
procedures is presented in Appendix F. Calibration data
is included in Appendix G.
PARTICULATE SAMPLING
Prior to particulate sampling..of the waste heat
exhaust stack, preliminary determinations were made
to select a sampling nozzle of proper size to maintain
isokinetic sampling rates throughout the sampling
study. During these preliminary-determinations the
following were executed:
(1) The establishment of a minimum number
of sampling points for the sampling site,
calculated according to Method 1;
(2) The measurements of the stack static
pressure, stack velocity pressure
profile and temperatures at each sampling
point, per Method 2 procedures;
(3) Determination of the stack gas dry
molecular weight as described in Method 3;
and,
(4) the approximate stack gas moisture content
using Method 4.
While conducting the moisture test, the metering
system (i.e., the vacuum pump vanes) malfunctioned
- 27 -
-------
and was unable to maintain a steady flowrate for the
remainder of the test. This preliminary test resulted
in a moisture value of 3-percent, which was 10-
percent lower than the average moisture value from a
previous testing program for this source. The approx-
•*•
imate moisture content thus determined was deduced
to be suspect. Therefore, the average moisture value
(13-percent) from previous battery stack emission studies
was used for the preliminary determination of the
isokinetic sampling rate. This assumed moisture value
turned out to be 0.9-percent below the average actual
percent moisture data for this emission study.
Triplicate 144-minute particulate samples were
extracted isokinetically for 4^minutes at each of 32
of the 36 sampling points in the waste heat stack.
Points 1 and 18 on each traverse diameter were judged
to be too close to the sampling port liners, which were
not flush with the stack wall, so points 2 and 17
were sampled for twice the normal duration, thus deleting
four sampling points.
As it would have been difficult to support and move
an impinger box at the end of a 16-foot probe it was t
necessary to modify an EPA Method 5 particulate sampling
train. The heated filter was placed at the end of the
probe and connected to the impingers with a flexible
Teflon® line (Figure 5.1.). The sampling train consisted
- 28 -
-------
Unheated
stainless
steel probe
H eat ed 110-mm
Type A glas s-
fiber filter
ro
vo
Braided Teflon
tub ing
S-type Pitot
tub e
Pyrometer
100-ml
distilled
water
Thermometers
Micromanometer
Dry gas
meter
Main Vacuum
valve gauge
Inclined
manometer
Vacuum
pump
Figure 5.1. Particulate sampling train.
-------
of (in sequential order of sampled gas flow): a sharp,
tapered, stainless steel (SS) sampling nozzle; a 16-
foot SS probe assembly (instead of a glass probe, due
to the obvious probability of breakage during testing);
a heated, pre-weighed 110- millimeter (mm) Type-A
"<§)
glass-fiber filter; flexible Teflon^ tubing leading to
two Greenburg-Smith impingers, the first modified, the
second standard, each containing 100-mil1i1iters (mis)
of distilled water; an empty modified Greenburg-Smith
impinger; a modified Greenburg-Smith impinger contain-
ing approximately 400 grams of silica gel; a leakless
pump with vacuum gauge; a calibrated dry-gas meter
equipped with bimetallic inlet and outlet thermometers;
and, a calibrated orifice-type flowmeter connected ,'
to a zero to ten-inch range inclined (water gauge)
manometer. j
The impinger train was immersed in an ice bath
to maintain the temperature in the last impinger
at 70F or less. A calibrated S-type Pitot tube
was connected to the sampling probe and velocity
pressures were read on a zero to two-inch micro-
manometer. An iron-constantan (I/C) thermocouple,
attached to the Pitot-probe assembly, was connected
to a calibrated pyrometer. During the course of testing,
- 30 -
-------
the probe and filter temperatures were kept above the
dew point of the exhaust gases sampled.
The sampling train was checked for leaks before
and after each sample run in accordance with the
requirement that the initial leak rate shall not exceed
*•
0.02 cfm at 15-inches of mercury vacuum and the final leak
rate shall not exceed 0.02 cfm at the greatest
vacuum which occurred during the test.
During each test, the probe, Pitot tube and
thermocouple assembly were moved to each sampling
point, the velocity pressure and temperature of
the exhaust gas were measured, and isokinetic
sampling flowrates were adjusted accordingly,
\
using an orifice-type meter to indicate instantaneous
flowrates.
Following the leak check at the end of each
144-minute sample run, the sampling train was
transferred to a sheltered clean-up area. The
volumes of the impinger contents were measured
and volume increases recorded. The solutions were
placed in glass sample bottles and sealed with
Teflon®-lined caps. The silica gel was weighed to
determine the weight gain (as condensate). The
probe and nozzle assembly was initially rinsed
and brushed with water and then with acetone.
- 31 -
-------
The two rinsings were collected in separate glass
sample bottles with Teflon®-lined caps. The impinger
assembly was thoroughly rinsed with water, and these
water rinsings were placed with the impinger solutions,
Following the water wash of the impingers, the entire
impinger assembly was then rinsed with acetone and
these rinsings were placed in separate glass sample
bottles and sealed with Teflon®-lined caps. The
glass-fiber filter was carefully returned to its
original petri dish and sealed for transport.
The front half of the filter holder was rinsed
and brushed with water and acetone and these
rinsings were added to the probe rinses. The back
half of the filter holder and the Teflon® flexline
were rinsed and brushed with water and acetone,
and these rinsings were added to the contents and
rinsings of the impingers. Thus, five fractions
were collected from each sample run:
(1) water washings of nozzle, probe and front
half of filter holder;
(2) acetone washings of nozzle, probe and front
half of filter holder;
(3) 110-mm Type-A glass-fiber filter;
- 32 -
-------
(4) irapinger contents and distilled water rinsings
of back half of filter holder, Teflon® flex-
line and impingers; and,
(5) acetone rinsings of back half of filter holder,
Teflon® flexline and impingers.
In the laboratory, all liquid" fractions were observed
for leakage, then each measured volumetrically and the
values recorded.
PARTICULATE EMISSIONS
A 50-ml aliquot was removed from Fractions 1 and 4
prior to the particulate analyses for sulfate analysis.
The Fraction 1 samples were then transferred to tared
beakers and evaporated to dryness at 105C. The dried
residues were desiccated for at least 24 hours before
determining constant weights. (A constant weight is
determined by redesiccating the sample for at least 6-
hours and then reweighing it. A difference between
these weights less than 0.5 milligram, or 1-percent
of the total weight, constitutes a constant weight.)
The acetone volumes of Fractions 2 and 5 were transferred
to tared beakers, evaporated to a residue at ambient
conditions, then desiccated for 24-hours before
constant weights were determined. The 110-mm
Type A filters were desiccated at ambient conditions for
more than 24 hours to determine constant weights.
- 33 -
-------
The particulate and sulfate weights by fraction are
presented in Appendix H.
Chloroform/Ether Extraction
After sulfate aliquots (50-ral) were removed,
Fraction 4 water samples were extracted three times
with equal volumes (30-ml) of chloroform and ethyl
ether (C/E). The organic phases (C/E extractables)
of each sample were collected and combined in separate
tared beakers, evaporated to dryness at ambient
conditions, and desiccated for 24 hours before constant
weights were determined. The remaining water phase
of Fraction 4 samples were evaporated to dryness at
105C, desiccated 24 hours and measured to constant
weights. The sum of the C/E extractable organic
weight and the aqueous phase weight (inorganic
fraction) is the total weight of Fraction 4.
Sulfate Analysis
The 110-mm filters (Fraction 3) were leached of
sulfates by liquefying each filter with 80-percent
isopropanol (IPA) in a blender, scrubbing the amalgamated
solutions in an ultrasonic bath, punctually followed
by filtering the solutions and diluting the filtrate
to 150 mis with 80-percent IPA. The residues from Fractions
2 and 5 were leached of sulfates by scrubbing the residue
with 80-percent IPA in an ultrasonic bath then bringing
to 100-ml volume with 80-percent IPA,, Five-mi portions
- 34 -
-------
of 50-ml aliquots of Fractions 1 and 4 were brought
to 25-ml volumes with 100-percent isopropanol to
form 80-percent solutions.
Each of the above solutions was adjusted for
acidity with perchloric acid to a pH between 2.5
and 4.0. Three to five drops of thorin indicator
were added to each solution before titrating with
standardized barium perchlorate to a pink endpoint.
The results are reported as sulfuric acid (including
sulfur trioxide) and as a percent of the particulate
weight for the individual fractions.
Exhaust Gas Composition
An integrated gaseous sample was withdrawn
simultaneously with each particulate sample from
the south port of the waste heat stack. The Orsat
sampling train utilized during the first two sample
runs consisted of a SS probe; a particulate/condensate
trap; a leakless diaphragm pump; a pressure release
connection; a needle valve coupled with a rotameter;
and a 96-liter Saran® sample bag. All intermediate
connections were made with lengths of polyvinylchloride
(PVC) tubing. See Figure 5.2 for a graphic display
of the sampling train used in gaseous sampling.
The Orsat sampling train was modified for
Run 3, in that the carbon monoxide (CO) sampling
- 35 -
-------
I
U)
Stainless steel probe
PVC tubing
Particulate/
Condensate trap
Diaphragm
vacuum
pump
R.o tame ter-
Pressure relief valve
96 Liter
Saran gas bag
\
Figure 5.2, Orsat gas sampling train
-------
probe, particulate/condensate trap, and intermediate
Teflon® tubing replaced their counterparts from
the previous Orsat train. In addition,a 3-way brass
valve was inserted after the particulate/condensate trap
**
to split the gaseous sample for both continuous CO
monitoring and integrated bag sample collection.
Each integrated bag sample was analyzed by the
Orsat method for carbon dioxide, oxygen and carbon
monoxide concentrations as specified in EPA Method 3.
The results were used to calculate the molecular
weight of the waste heat exhaust gas in the Battery
D stack.
CARBON MONOXIDE SAMPLING
A sample of flue gas was drawn through a SS
probe, Teflon^ tubing and a particulate/condensate
trap containing glass wool, to a pair of modified
Greenburg-Smith impingers. The first impinger
contained approximately 250 grams of silica gel
(R)
and the second approximately 500 grams of Ascarite^y
to remove moisture and carbon dioxide, respectively.
Finally, a leak-free diaphragm pump forced the sample
through a needle valve and rotameter to a Beckman,
Model 865, NDIR Analyzer.
The sampling probe was positioned in the west
port of the stack for the first and second particulate
runs but was moved to the south port for Run 3.
- 37 -
-------
The sampling train was modified for the third sample
run to accommodate the simultaneous collection of
an integrated bag sample for the determination of
exhaust gas composition by the Orsat method. The
modification consisted of inserting a 3-way brass
*•
valve after the particulate/condensate trap, which
was used to divide the gas sample into two streams,
one for the Beckman NDIR Analyzer and the other for
the Orsat integrated gas sample.
At the sample interface to the Beckman NDIR
Analyzer, an approximate flowrate and delivery pressure
of 1.5 cfh and 10 psig respectively, were maintained
for the duration of continuous sampling. An analog
strip chart recorder was used to record all instrument
outputs. This sampling system is depicted in Figure
5.3.
The daily calibration sequence included passing
a certified standard zero gas (dry nitrogen) and a
certified standard span gas concentration (carbon
monoxide in nitrogen) through the analyzer. The
instrument output was calibrated for the anticipated
range of 0 to 10,000 ppm carbon monoxide by adjustment
of the zero and gain levels to the appropriate signal,
as indicated on a calibration curve. Following
the analyzer calibration, the strip chart recorder
pen response was increased by a factor of ten by
changing its input range.
- 38 -
-------
Stainless steel probe
Teflon sampling line
Three-way valve
To integrated
bag sample
(for Run 3 only)
(jO
\0
250g 500g
silica Ascarite^
gel
Needle
va Ive
Calibration
gases
Beckman Model
865 NDIR
analyzer
Flowrate
Meter
Ranges(°-10>000
(0-30,000 ppm
Figure 5.3. Sampling train for continuous monitoring of carbon monoxide.
-------
The Beckman NDIR analyzer, which operates by the
Luft principle specified by Reference Method 10, was
equipped with a four-position valve to al-low the
introduction of sample gas or any of the required
standard calibration gases, as depicted in Figure
•*•
5.3.
The actual measured concentrations of carbon
monoxide were determined by adjusting the recorded
strip chart values with the factory calibration curve
for the 16-mm cell, which had been adjusted to the
standard gas concentrations used in the field.
COMBINATION SAMPLE EXTRACTION SYSTEM FOR NITROGEN
OXIDES AND OXYGEN
An in-stack stainless steel (SS) alundum thimble
holder, packed with glass wool, was positioned in the
North port of the D Battery stack throughout the complete
testing study. The intermediate connections between
components were accomplished using lengths of Teflot®
tubing. A three-way SS valve was incorporated in-line
after the probe to branch off to a leakless diaphragm
pump. The 3-way valve facilitated sampling system
leak checks and back purging of the probe assembly.
Subsequently, the sampled gas flowed into an ice-
cooled condenser and through a leakless diaphragm pump.
- 40 -
-------
A SS control valve maintained the gas flow at 2 4-
0.5 scfh at the outlet of the rotameter downstream
of both analyzers. A pair of SS "tee" connectors
allowed aliquots of gases to be bled off to the
analyzers. .See Figure 5.4 for the schematic of
.*•
the continuous extraction system and NOX-C>2 analyzers.
A leak check of the system was performed prior to
each day's testing by closing the 3-way valve following
the probe assembly and observing the deflection of a
rotameter at the outlet of the sampling system. Any
noticeable deflection after a 2-minute time period
indicated a leak in the sampling system.
CONTINUOUS NITROGEN OXIDES MONITORING
An aliquot of exhaust gases was withdrawn from
the "tee" connector through a 3-way SS valve to a
Thermo-Electron NOx Analyzer. The sampled gas flow
through the NO^ analyzer was monitored and maintained
at 2.5 scfh by means of a flow meter mounted on the
panel of the analyzer. Ambient air (dried) was employed
as an oxygen source for the ozone generator within
the analyzer. The vacuum pressure to the reaction
chamber was monitored by a gauge connected with the
chamber unit. The analyzer output signal was connected
to a linear strip chart recorder which was set at a
chart speed of 0.5 centimeters/minute (cm/min).
- 41 -
-------
ss
Alundum
thimble
holder
Stack
wall
SS probe
3-way valve
Diaphragm
pump
Condens or
/ £r~>s
r^'
/Diaphragm pump.
To analyzers
From probe
ScS control
valve
\ 4V^
Calibra-
Calibration J. „ tion .
gas
Vacuum Q
gauge \—
I
/"^ — J
Diaphragm / , — — >
pump V^ix
(For ozone
generator)
;fcw__KJ-way . ^
•^T' valve. gas ^
T
i
^~P 3-way
valve
TECO® Teledyne®
®0X °2
analyzer analyzer
- , Linear . Linear
Instruments® Instruments®
Chart Recorder | Chart Recorder
I
i
6
Rotame ter
(To measure
exce s s flow)
Figure 5.4.
Combination sampling trains for continuous monitoring
of nitrogen oxides and oxygen.
- 42 -
-------
Calibration checks were performed prior to the start-
up of each run, at selected times during testing, and
at the completion of each day's testing,,
CONTINUOUS OXYGEN MONITORING
A separate aliquot of the exhaust gases was withdrawn
from the "tee" connector through a 3-way SS valve to a
Teledyne Model 326A Oxygen Analyzer.. The flowrate was
continually maintained at 2 to. 2.5 scfh throughout the
testing program. The analyzer output signal was
connected to a linear strip chart recorder with a chart
speed of one-inch per hour. Calibration checks were
routinely performed.
NITROGEN OXIDES GRAB SAMPLES
Daily "grab samples" were acquired from the
waste heat stack for nitrogen oxides analysis,according
to EPA Method 7. The nitrogen oxides sampling train
apparatus is depicted in Figure 5.5. The field data
for each sample is included in Appendix B-3.
After evacuating each flask containing dilute
sulfuric acid/hydrogen peroxide absorbing solution to
the appropriate vacuum pressure and leak checking,
the SS probe and glass stopcock were purged with
stack gas for approximately one minute. The stop-
cock was then returned to the sampling position for
- 43 -
-------
-p-
I
_^ Stainless steel probe
Mercury
s lack
tube
Figure 5.5. Sampling train for nitrogen oxides grab samples.
-------
a period of approximately thirty seconds to insure
equal pressurization between the sampling system and
stack gases. Immediately following sample collection,
each flask was shaken for five minutes and stored for
*•
at least 16 hours away from sunlight.
Prior to sample recovery, each flask was shaken for
two minutes, followed by the measuring and recording of
the internal pressure and temperature of the flask. The
contents of each flask were transferred to leak-free
polyethylene bottles along with two 5-ml water rinses.
The pH of each solution was adjusted to a range of
9 to 12 with l.ON. sodium hydroxide before the bottles
were marked, labelled and sealed for transport to
the laboratory.
In the laboratory, the sample bottles were
examined for leakage before transferring their
contents to tared beakers along with sample
bottle water rinsings. The samples were evaporated
to dryness and treated successively with solutions
of phenol disulfonic acid, distilled water, and
sulfuric acid. The resulting solution was made
basic with ammonium hydroxide, transferred to a
volumetric flask, and diluted to volume with
distilled water. The standard wavelength absorbance
at 410 nanometers was measured and recorded as
nitrogen dioxide in the samples and potassium nitrate
standards.
- -45 -
-------
VISIBLE EMISSIONS
Visible emissions from the D Battery stack exhaust
were recorded for the duration of each sample run
except for Run 3, for which visible emissions
recording was abbreviated due to foul weather. The
*•
observations were performed in accordance with EPA
Method 9 by a qualified visible emissions observer.
A summary of the visible emission data is presented
in Appendix B-4.
- 46 -
-------
SECTION II
-------
BaP SAMPLING AND ANALYSIS II
i
The sampling train used to collect BaP was identical to the participate
train except the BaP train contained an absorbent module between the filter
and the first impinger.
The module was packed with a polymeric absorbent, XAD-2 (styrene divinyl
benzene). The temperature of water circulating through the cooling jacket was
maintained at 127°F, therefore, the sampled gas would pass through the adsorbent
material at a constant cooled temperature. The adsorbent module was covered
with aluminum foil throughout the testing to prevent deterioration of the sample
from exposure to ultraviolet light. The described operation of the absorbent
module was the only variance of the BaP train from the particulate train.
The moisture content of the stack gas must take into account all the water
collected in the BaP train. The water collected in the impingers and silica
gel will not accurately reflect the true moisture content of the stack gas as
the adsorbent module is located immediately behind the heated filter and water
cooled to 127°F; this temperature change will cause condensation of the stack
gas in the middle before the impingers. Therefore, in operating a BaP train
at a source with a high moisture content, either a moisture train or a Method 5
train should.be operated during the run. This would provide an accurate moisture
determination for the BaP run. This test utilized the moisture content from
the Method 5 train as the source for BaP data reductions.
-------
CF & I Battery Outlet
RUN
DATE
Meter Volume (DSCFJ
Meter Volume (DSCM)
Stack Volume (ACFM)
Stack Volume (DSCFM)
Stack Volume (DSCMM)
% Isokinetic
mg/Filter
mg/Rinse
mg/XAD-2
mg/Total
mg/DSCM
kg/Hour
Ibs/DSCF
Ibs/Hour
lbs/24 Hour Day
lbs/365 Days
BaP -
1
8-7-79
46.26
1.31
22,060
10,032
284.1
104.0
0.350
11.288
0.725
12.363
0.0094
0.0002
58.68 x 10"1!
3.532 x 10,
8.48 x Itf3
3.10
Test Results
2
8-8-79
95.34
2.70
19,280 19
8,460 8
239.6
125.1
0.350
7.236
0.300
7.886
0.0029
0.00004
18.10 x 10'1!
9.188 x TO,
2.20 x 10'-3
' 0.80
3
8-9-79
81.92
2.32
,976
,820
249.8
108.8
0.200
11.618
1.200
13.018
0.0056
0.00008
34.96 x 10'1]
1.850 x 10,
4.440 x ]Q~*
1.62
Average
74.51
2.11
20,438
9,104
257.8
112.6
0.300
10.047
0.742
11.089
0.0060
0.00009
37.45 x 10~
2.046 x 10
4.910 x 10
1.79
------- |