United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 80-IBR-12
Air
&EPA
industrial Boiiers
Emission Test Report
Du Pont Corporation
Parkersburg, West Virginia
-------
INDUSTRIAL BOILERS
Emission Test Report
Du Pont Corporation
Parkersburg, West Virginia
December 15-17, 1980
Technical Directive No. 6
Prepared for
Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Measurement Branch
Research Triangle Park
North Carolina 27711
by
C. L. Cornett, Jr. and R. G. Beer
Contract No. 68-02-3547, Work Assignment No. 2
(EMB 80-IBR-12) (ESED 76/13)
February 1982
MONSANTO RESEARCH CORPORATION
DAYTON LABORATORY
Dayton, Ohio 45407
-------
CONTENTS
Page
Figures v
Tables vi
1. Introduction 1
2. Summary of Results 3
Description of Monitoring 3
Test Results 5
3. Process Description and Operation 25
Coal Handling System 25
Boiler Operation 26
Flue Gas/Air Pollution Control System 29
Ash Handling System 30
Operating Conditions 31
4. Location of Sampling Points 39
Control System Inlet (North Side) 39
Control System Inlet (South Side) 44
Control System Exhaust 44
Coal Sampling Site 47
5. Description of Sampling Trains 48
Particulate Sampling Train 48
Particulate Si2e Sampling Train 48
6. Sampling and Analytical Procedures 52
Summary 52
Field Sampling 52
Sample Handling 53
Chemical Analysis 53
Coal Analysis 55
Data Reduction 56
Quality Assurance 1 .... 57
111
-------
CONTENTS (continued)
Appendices
A. Field Sampling Data Sheets and Computer Coding
Forms A-l
B. Printouts of Sampling Results B-l
C. Visible Emission Field Data Sheets C-l
D. Boiler Monitoring Data D-l
E. Analytical Methods for Sulfuric Acid and
Sulfate Determinations E-l
F. Analytical Data F-l
G. Project Participants G-l
IV
-------
FIGURES
Number Page
1 Cumulative size distribution for north side inlet
Andersen run (run no. M5B-I-R2) at Du Pont Boiler
No. 3 in Parkersburg, WV (December 17, 1980) .... 14
2 Cumulative size distribution for south side inlet
Andersen run (run no. M5B-I-R1) at Du Pont Boiler
No. 3 in Parkersburg, WV (December 17, 1980) .... 15
3 Cumulative size distribution for outlet Andersen run
(run no. M5B-0-R1) at Du Pont Boiler No. 3 in Par-
kersburg, WV (December 17, 1980) 16
4 Process flow schematic of Du Pont Washington Works
Boiler No. 3 27
5 Schematic diagram of Du Pont Boiler No. 3 and
associated equipment in Parkersburg, WV 40
6 Top view of inlet sampling area at Du Pont
Boiler No. 3 in Parkersburg, WV 41
7 North side inlet sampling site at Du Pont
Parkersburg, WV Boiler No. 3 42
8 Inside view of inlet duct at Du Pont Boiler No. 3
in Parkersburg, WV 43
9 South side inlet sampling site at Du Pont
Parkersburg, WV Boiler No. 3 45
10 Exhaust sampling site at Du Pont Parkersburg, WV
Boiler No. 3 46
11 Dual-probe sampling train 50
12 Particle size distribution sampling apparatus -
Andersen 2000, Inc 51
-------
TABLES
Number Page
1 Source Sampling and Analysis at Du Pont
Boiler No. 3 in Parkersburg, WV
(December 16-17, 1980) 4
2 Emission Data for Du Pont Boiler No. 3 in
Parkersburg, WV, December 16-17, 1980
(Metric Units) 6
3 Emission Data for Du Pont Boiler No. 3 in
Parkersburg, WV, December 16-17, 1980
(English Units) .... 7
4 Summary of Mechanical Collector Efficiency Measure-
ments and Differences Between Method 5 and Method 5B
Results at Du Pont Boiler No. 3 in Parkersburg,
WV (December 16-17, 1980) 9
5 Summary of Duration of Sampling, Stack Temperature,
Stack Flow Rate, Sample Volume, and Sample Water
Content at Du Pont Boiler No. 3 in
Parkersburg, WV (December 16-17, 1980) 10
6 Summary of Integrated Gas Analysis Results at
Du Pont Boiler No. 3 in Parkersburg, WV
(December 16-17, 1980) 12
7 Summary of Andersen Particle Size Results at
Du Pont Boiler No. 3 in Parkersburg, WV
(December 17, 1980) 13
8a Summary of Method 9 Plume Opacity Observations at
Du Pont Boiler No. 3 in Parkersburg, WV (Decem-
ber 16, 1980) 17
8b Summary of Plume Opacity Observations at Du Pont
Boiler No. 3 in Parkersburg, WV
(December 17, 1980) 19
9 Summary of Coal Analysis at Du Pont Boiler No. 3
in Parkersburg, WV (December 16-17, 1980) 22
10 Summary of Process Conditions During Testing at
Du Pont Boiler No. 3 in Parkersburg, WV
(December 16-17, 1980) 32
VI
-------
SECTION 1
INTRODUCTION
Boiler No. 3 at the Du Pont Washington Works on Du Pont Road in
Parkersburg, West Virginia was tested for particulate, sulfuric
acid and sulfate emissions on December 16-17, 1980, by Monsanto
Research Corporation (MRC). This work was performed for the
Emission Measurement Branch of the U.S. Environmental Protection
Agency (EPA) under Contract No. 68-02-3547, Work Assignment
No. 2. The boiler tested was a 9.5 kg/s (75,000 Ib/hr) steam
capacity waste oil and coal-fired spreader stoker with an econo-
mizer and a two-stage mechanical collector for air pollution
control. Waste oil was not fired during the test.
The purpose of the sampling program is to provide background
information on well-controlled industrial boilers for the devel-
opment of new source performance standards. Within this frame-
work, the objectives of the sampling at the Du Pont Washington
Works were:
• To determine the effect of raising the temperature of the
filter and probe on an EPA Method 5 train from 120°C (248°F)
to 177°C (350°F) on the amounts of particulate, sulfate, and
sulfuric acid emissions measured downstream of the two-stage
mechanical collector (Method 5 testing, with the filter and
probe at 177 ± 14°C [350 ± 25°F] is called Method 5B testing);
• To determine the effect of the two-stage mechanical collec-
tor on particulate, sulfate, and sulfuric acid emissions,
at boiler loadings in excess of 90 percent capacity based on
-------
three EPA Method 5B test runs and at a reduced boiler load
of approximately 60 percent (or the minimum achievable
loading) based on a single EPA Method 5B test run.
-------
SECTION 2
SUMMARY OF RESULTS
DESCRIPTION OF MONITORING
Table 1 summarizes the monitoring that was performed at this
plant. It consisted of four test runs of Methods 5 and 5B at the
mechanical collector exhaust performed simultaneously with Meth-
od 5B sampling at the mechanical collector inlet. A special dual
probe stack test system was used for sampling with Methods 5 and
5B simultaneously through the same 4-inch ports at the exhaust,
with the centers of the nozzles spaced 1.75 in. apart. Sample
volumes of at least 1.25 dry standard m3 (44 dry standard ft3)
were taken during these runs. The boiler was running under
steady state conditions (no soot blowing, ash unloading, etc.) in
the 92 to 100 percent capacity range during the first three runs.
The fourth run was conducted at a 75 to 80 percent production
rate, which was the minimum achievable boiler loading allowed by
the plant steam demand.
The filters and dried acetone washes of the Method 5 and SB runs
were weighed and analyzed for sulfuric acid and sulfates.
Methods 1 through 4 were used during all Method 5 and 5B sampling
runs, as in typical compliance monitoring.
Particle size distributions were measured at full production
rates using an Anderson cascade impactor. Samples were collected
twice at the inlet to the two-stage mechanical collector and once
at the outlet. In addition, the plume opacity was measured
during all emission tests by EPA Method 9, when possible.
-------
TABLE 1. SOURCE SAMPLING AND ANALYSIS AT DU PONT BOILER NO. 3
IN PARKERSBURG, WV (DECEMBER 16-17, 1980)
SAMPLING AMD MM.VSIS MQUIROWTS
*t Job, ft. ffl
Contract Ho. 68-02-BSlTl*"**"*"1 Ha**r: ti.
»«»"««1
Coaptny NIM:
pont
Industry:
INDUSTRIAL BOILER
IEo^uny Location: Parkersburg, W. Va
--• | ' —"
Proc«»: Coal-Fired
Spreader Stoker
Control C«i1p«at: 2 Stage
Mechanical Collector
ToUl
no. of
(Mplll
$**>!•
Mtftod
SMOlt
colltcttd
fcy
Ntnlwi
tMpMng
tlM
HlntBMi
VOlMI 9<>
UBDltd ft*
InltUl Anilyili
Typt Mtlwd By
FtMl
Typt
Hithod ly
MKTICI/UTI
'HOKTH llJi.tr
1-1
vua.ni itm
ruin ".
1-1
1-1
10 HSO
nmii/urrrtaom
S'
1-1
±f*-
AriTl
uiotueu
VCKA/IX OIW
8mfcs
2-1
inTrfjA/r/t
HHtfHtfjt
2-1
tMDUUM
tortiruaur
nor unmet
COAL
at it
Utrf
.('^ ^
Twelve samples were combined into aggregate samples for each of four runs of Method 5a
at the inlet and Methods 5 and 5a at the exhaust. The 13th sample was associated with
the Andersen runs at the inlet.
-------
Feed coal samples were taken at the spreader-stoker inlet to the
boiler. Samples were composited from individual samples col-
lected before, during, and after each test run. The composited
samples were analyzed using ASTM ultimate analysis procedures.
TEST RESULTS
Particulate, sulfuric acid, and sulfate emissions measured at
both the collector inlet and outlet are summarized in Table 2 and
3, in metric and english units, respectively. Particulate emis-
sions represent all emissions, including sulfuric acid and sul-
fates, as measured by weighing the method 5 and 5B samples. The
emissions are presented as concentrations in the flue gas, mass
flow rates, and mass rates normalized to the boiler heat input of
the coal. The results of the fourth test run, at the 75 to
80 percent production rate, show the particulate concentrations
and normalized emission rates at only 50 to 60 percent of emis-
sions during the first three runs. The actual particulate con-
centrations to an equivalent 12 percent C02 flue gas content
basis is also presented in both tables.
In addition, Table 3 presents two extra columns. One column
shows that the percent isokinetics during all sampling was within
the acceptable range of 90 to 110. The last column shows the
percent of the concentrations measured by Method 5 represented by
the concentrations measured by Method 5B simultaneously at the
same location. Particulate concentrations measured by Method 5B
were 66% to 68% of the Method 5 particulate concentrations.
Sulfuric acid concentrations measured by Method 5B were 7.5% to
12.3% of the Method 5 sulfuric acid concentrations. Sulfate
concentrations measured by Method 5B were 70% to 91% of the
Method 5 sulfate concentrations. Therefore, the higher probe
temperature used in Method 5B minimizes the amount of sulfur
compounds caught in the probe and on the sample filter. The
results of Method 5B sampling are thus better representations of
sulfur-free particulate emissions.
5
-------
TABLE 2. EMISSION DATA FOR DU PONT BOILER NO. 3
IN PARKERSBURG, WV, DECEMBER 16-17, 1980
(METRIC UNITS)
Emissions
Run
number Date Location
1 12-16-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
2 12-16-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
3 12-17-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
4C 12-17-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Sampling
method Pollutant
SB
5
SB
SB
5
SB
SB
S
SB
SB
5
SB
SB
5
SB
SB
5
SB
SB
5
SB
SB
5
SB
SB
S
SB
SB
5
SB
SB
5
SB
SB
5
SB
P articulate
Farticulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sul fates
Sul fates
Sul fates
Particulate
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sul fates
Sulfates
Sul fates
Particulate
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Sulfates
Particulate
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Sulfates
Actual
q/dscm
3
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
.0068b
.3908
.2674
.0165b
.0920
.0069
.0808b
.0468
.0324
.1627
.3520
.2340
.0095
.0692
.0084
.0484
.0445
.0322
.9980
.3471
.2323
.0109
.1162
.0101
.0652
.0350
.0296
.3645
.2056
.1370
.0260
.0864
.0099
.0798
.0194
.0176
kd/hr
138
19
13
0
4
0
3
2
1
108
16
11
0
3
0
2
2
1
117
17
11
0
5
0
2
1
1
80
7
5
0
3
0
2
0
0
.5b
.4
.3
.8b
.6
.3
.7b
.3
.6
.7
.9
.3
.5
.3
.4
.4
.1
.6
.4
.7
.8
.4
.9
.5
.6
.8
.5
.4
.7
.2
.9
.2
.4
.7
.7
.7
n
-------
TABLE 3. EMISSION DATA FOR DU PONT BOILER NO. 3 IN PARKERSBURG,
WV, DECEMBER 16-17, 1980 (ENGLISH UNITS)
Emissions
Run
number
1
2
3
4C
"This is
tration
Because
Sampling
Date Location method Pollutant
12-16-80
12-16-80
12-17-80
12-17-80
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
SB
S
SB
SB
S .
SB
5B
5
SB
SB
S
SB
SB
5
SB
SB
5
SB
SB
S
SB
SB
S
SB
SB
S
SB
SB
5
SB
SB
5
SB
SB
S
SB
Parti culate
P articulate
Farticulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Sulfates
Particulate
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sul fates
Sulfates
Particulate
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Sulfates
Particulate
Particulate
Farticulate
Sulfuric acid
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Sulfates
Actual
gr/dscf
1.3137"
0.1707
0.1168
0.0072b
0.0402
0.0030
0.0353b
0.0204
0.0142
0.9448
0.1538
0.1022
0.0042
0.0302
0.0037
0.0211
0.0194
0.0141
1.3098
0.1516
0.1015
0.0048
0.0507
0.0044
0.0285
0.0153
0.0129
1.0330.
0.0898?
0.0599
0.0113^
0.0377°
0.0043
0.0349
0.0085
0.0077
Corrected ,
to 12X COz,
Ib/hr
305. 4b
42.8
29.3
1.7"
10.1
0.8
e.2b
S.I
3.6
239.7
37.4
24.9
1.1
7.3
0.9
5.4
4.7
3.4
258.8
38.9
26.0
0.9
13.0
1.1
S.6
3.9
3.3
177. «.
17. Od
11. 4d
l-9A
7.1d
0.8d
6.0
1.6
1.5
the concentration normalized to 12% CO, . C = C x eMr. where C is
in the stack, and % CO. is the X CO, measured in tfl« stack.
the integrated gas
CO] observed in the
samples
sampler
at the
on the inlet was found to leak
control system exhaui
Run 4 was at a reduced production rate compared to runs
Because
the integrated gas
CO2 observed in the
samples
sampler
at the
Ib/mm Btu
3.891b
0.506
0.346
0.021b
0.119
0.009
0.105b
0.060
0.042
3. 259
0.392
0.261
0.014
0.077
0.009
0.073
0.049
0.036
3.959
0.494
0.331
0.015
0.165
0.014
0.086
0.050
0.042
2.912,,
0.253d
0.169d
0.032.
0.106?
0.012d
0.098
0.024
0.022
the measured
after the sample was
it were used in the
1, 2, and
3.
on the exhaust was found to leak after
control system inlet
were used
or/dscf
1.8331b
0.2382
0.1630
0.0100b
0.0561
0 . 0042
0.0493b'
0.0285
0.0198
1.61*7
0.1883
0.1251
0.0072
0.0370
0 . 0045
0.0362
0.0238
0.0173
1.9168
0.2303
0.1542
0.0070
0.0770
0.0067
0.0417
0.0232
0.0196
1.4086d
0.1225d
0.0817°
0.0154d
0.0514?
0.0059
0.0476
0.0116
0.0105
concen-
taken, the
Percent of
Method 5
gr/dscf
represented by
Method SB
Percent
gr/dscf
isokinetic at outlet
101.1
103.0
106.4
101.1
103.0
106.4
101.1
103.0
106.4
98.4
104.3
106.0
98.4
104.3
106.0
98.4
104.3
106.0
102.8
103.0
103.9
102.3
103.0
103.9
102.3
103.0
103.9
102.4
107.2
101.1
102.4
107.2
101.1
102.4
107.2
101.1
68
7
69
66
12
72
67
8
84
66
11
90
.4
.S
.6
.4
.3
.7
.0
.7
.3
.7
.4
.6
percents oxygen and
inlet calculations.
the sample was taken, the percent*
oxygen and
in the exhaust calculations.
-------
Appendix A-3 shows the weights of particulate, sulfuric acid, and
sulfates in the filters, and acetone rinses separately. Computer
printouts of the particulate, sulfuric acid, and sulfate emission
rates are provided in Appendix B.
Table 4 summarizes the particulate, sulfate, and sulfuric acid
collection efficiency of the two-stage mechanical collector, as
measured by EPA Method 5B. The differences between emissions
measured by Method 5 and 5B in the mechanical collector exhaust
are summarized in Table 4. Table 4 also summarizes what percent
of the differences between particulate emissions measured by Meth-
od 5 and 5B can be accounted for by differences in the amounts of
sulfuric acid (and sulfuric acid and sulfates combined) measured
by Method 5 and 5B. Fifty-one to 112% of the amount of differ-
ence in particulate emissions measured by Method 5 versus 5B are
accounted for by differences in the amounts of measured sulfuric
acid emissions. Sulfuric acid and sulfates combined account for
62% to 114% of the difference in particulate emissions between
measurements by Method 5 and Method 5B. Collection efficiencies
for sulfates, sulfuric acid, and sulfuric acid and sulfate, was
based on concentrations determined by Method 5B sampling.
Table 5 summarizes the duration of the Method 5 and 5B sampling,
the sample volumes, the stack temperatures, the stack flow rates,
the water content and the stack static pressure of the flue gas
at the inlet and exhaust of the mechanical collector. The higher
flue gas flow rates at the pollution control device outlet com-
pared to the inlet (33-196 dscm/min) and the associated tempera-
ture drops (26-29°C) suggest air leakage at or around the pollu-
tion control equipment. Potential causes for this leakage include
unseated dust hopper valves, collector inspection door seal
leakage, etc. (Leakage from ports was minimized by covering the
ports with rags during sampling.) Excessive air infiltration,
especially at the pollution control equipment, could tend to
promote higher fly ash re-entrainment from the collector and
8
-------
TABLE 4. SUMMARY OF MECHANICAL COLLECTOR EFFICIENCY MEASUREMENTS AND DIFFERENCES
BETWEEN METHOD 5 AND METHOD 5B RESULTS AT DU PONT BOILER NO. 3
IN PARKERSBURG, WV (DECEMBER 16-17, 1980)
Run
number
1
2
3
4C
Pollutant
Particulate
Sulfuric acid
Sul fates
Sulfuric acid and
sul fates
Particulate
Sulfuric acid
Sul fates
Sulfuric acid and
sul fates
Particulate
Sulfuric acid
Sul fates
Sulfuric acid and
sul fates
Particulate
Sulfuric acid
Sul fates
Sulfuric acid and
sul fates
Collection
efficiency, %
Based
on kg/hr
(lb/hr)a
emissions
90.4
53
56
56
89.6
20
37
34
90.0
-20
40
32
93. 6b
"E
75b
71b
Percent of difference in particu-
Method 5 late emissions, kg/hr (Ib/hr)
minus Represented by
Method 5B . difference in
emissions sulfuric acid and
kg/hr
6.1
4.3
0.7
5.0
5.6
2.9
0.5
3.4
5.9
5.4
0.3
5.7
2.5b
2.8b
o.ob
2.8b
Ib/hr sul fates combined
13.5
9.3
1.5
10.8 80.0
12.5
6.4
1.3
7.7 61.6
12.9
11.9
0.6
12.5 96.9
5.6b
6.3?
O.lb
6.4b 114.2
Represented by
difference in
sulfuric acid
emissions
68.9
51.2
92.2
112.5
Collection efficiency for sulfuric acid, sulfates, and sulfuric acid and sulfates based on con-
centrations determined by Method 5B sampling.
"Because the integrated gas sampler on the exhaust was found to leak after the sample was taken,
the percents oxygen and CO2 observed in the samples at the control system inlet were used in
the exhaust calculations.
"Run 4 was at a reduced production rate compared to Runs 1, 2, and 3.
-------
TABLE 5. SUMMARY OF DURATION OF SAMPLING, STACK TEMPERATURE, STACK FLOW
RATE, SAMPLE VOLUME, SAMPLE WATER CONTENT, AND STATIC PRESSURE
AT DU PONT BOILER NO. 3 IN PARKERSBURG, WV (DECEMBER 16-17, 1980)
Run
No.
1
2
3
4
Location
Total inlet
Outlet
Outlet
Total inlet
Outlet
Outlet
Total inlet
Outlet
Outlet
Total inlet
Outlet
Outlet
Sampling
method
5B
5
SB
5B
5
SB
SB
5
SB
SB
5
SB
Duration
of
sampling,
min
144
144
144
144
144
144
144
144
144
144
144
144
Measured
stack
temperature
°C °F
182
156
156
182
155
156
184
156
156
175
146
146
360
313
313
359
312
313
363
313
313
347
295
295
Stack
flow rate
dscm/
min
768
829
830
838
803
805
653
849
848
567
626
630
dscf/
min
27,126
29,257
29,308
29,600
28,343
28,425
23,053
29,964
29,939
20,037
22,104
22,260
Sample
dscm
3.04
1.59
1.75
3.23
1.56
1.69
2.63
1.63
1.74
2.29
1.25
1.26
volume
dscf
107.37
56.01
61.80
114.03
54.92
59.68
92.84
57.39
61.60
80.79
44.03
44.59
Sample
water
content,
5.39
4.74
4.54
3.53
5.54
5.15
5.02
5.03
5.13
5.52
5.36
4.80
Stack
static
pressure,
in. H2O
-4.6
-0.02
-0.02
-4.6
-0.02
-0.02
-4.6
-0.02
-0.02
-4.6
-0.02
-0.02
-------
higher dust loadings causing the observed emission rates to be
higher than expected of a well sealed system. Lower air flow
rates were observed during run 4 because the boiler was operating
at a reduced capacity.
Table 6 summarizes the flue gas carbon dioxide, carbon monoxide,
oxygen, and nitrogen measurements associated with selected Meth-
od 5 and 5B samples. These values were used to calculate emis-
sions rates in Tables 2 and 3. Leaks in the integrated gas
sampling train prevented valid samples from being taken at the
inlet to the collector during the first run and at the exhaust
during the fourth run. When such data were unavailable, inlet
data were used for the missing exhaust data (and vice versa) when
the Method 5 and 5B data were reduced.
Table 7 summarizes the results of the Andersen cascade impactor
sampling. The cumulative aerodynamic size distributions eire
graphed in Figures 1, 2, and 3.
Table 8 summarizes the results of the plume opacity observations
by EPA Method 9. Opacity readings were not taken during the
second run because of darkness.
Table 9 summarizes the results of the coal sampling and analysis.
(According to plant personnel, the results are representative of
the coal feed normally used.) To obtain an indication of the
precision of the coal analysis data, half of a raw coal sample
that was previously analyzed (as part of a November 1980 emission
test at the General Motors Chevrolet Plant in Parma, Ohio) was
analyzed along with the Parkersburg samples. The last column of
Table 9 shows the difference in the results between the analyses
of different halves of the same sample. The most significant
differences were in the reported sulfur content (1.27%, dry) and
in the ash content (4.13%, dry). Portions of the same half-
samples discussed above were then analyzed a second time; the
11
-------
TABLE 6. SUMMARY OF INTEGRATED GAS ANALYSIS RESULTS AT DU PONT
BOILER NO. 3 IN PARKERSBURG, WV (DECEMBER 16-17, 1980)
Run
number
la
2
3
4b
Location
Outlet
Outlet
North inlet
South inlet
Outlet
South inlet
Outlet
South inlet
Sampling
method
before
Method 3
train
5
5B
5B
5B
5B
5B
5B
5B
Dry
molecular
C09
8
8
7
7
9
8
7
8
7
1 /O
.6
.6
.0
.0
.8
.2
.9
.8
CO,
0.
0.
0.
0.
0.
0.
0.
0.
0
0
0
0
0
0
0
0
o,. X
11.0
11.0
12.4
12.4
9.4
11.2
11.9
10.5
N?
80
80
80
80
80
80
80
80
weight,
kg/kg mole
, % (Ib/lb mole)
.4
.4
.6
.6
.8
.6
.2
.7
29
29
29
29
29
29
29
29
.81
.81
.61
.61
.94
.76
.75
.83
No inlet values are reported due to leak in sampler used at inlet.
No outlet values are reported due to leak insampler used at outlet.
-------
TABLE 7. SUMMARY OF ANDERSEN PARTICLE SIZE RESULTS AT DU PONT BOILER NO.
IN PARKERSBURG, WV (DECEMBER 17, 1980)
Flow rate
Location Run acm/min acf/min
North side
inlet M5B-I-R2 0.019 0.68
South side
inlet M5B-I-R1 0.018 0.62
Outlet M5B-0-R1 0.018 0.65
Percent
isokinetic Stage
103.7 0
1
2
3
4
5
6
7
Final filter
110 0
1
2
3
4
5
6
7
Final filter
105.4 0
1
2
3
4
5
6
7
Final filter
Size
range,
pm
>12.0
8.0-12.0
5.6-8.0
3.7-5.6
2.4-3.7
1.18-2.4
0.74-1.18
0.5-0.74
0-0.5
>12.8
8.6-12.8
5.8-8.6
3.95-5.8
2.5-3.95
1.25-2.5
0.79-1.25
0.54-0.79
0.0-0.54
>12.2
8.4-12.2
5.8-8.4
3.9-5.8
2.45-3.9
1.2-2.45
0.76-1.2
0.52-0.76
0-0.52
Percent
in size
range
90.98
0
2.63
0.68
1.03
0
0.11
0.57
4.00
97.52
0.13
0.13
0.13
0.38
0.31
0.13
0.13
1.14
37.16
5.75
1.15
0
16.86
13.31
8.81
1.15
16.09
Cumulative
percent,
-------
1.1
PAHTICLt: AERODYNAMIC" DIAMETER, Mm
Figure 1. Cumulative size distribution for north side inlet Andersen run
(run no. M5B-I-R2) at Du .Pont Boiler No. 3 in Parkersburg, WV
(December 17, 1980).
-------
tn
•.I
"•Vo ,10 j« yo f.o •• ii «»
PARTICLE AERODYNAMIC DIAMETER, urn
Figure 2. Cumulative size distribution for south side inlet Andersen run
(run no. M5B-I-R1) at Du Pont Boiler No. 3 in Parkersburg, WV
(December 17, 1980).
-------
1.1
Figure 3
PARTICLE AERODYNAMIC DIAMETER, inn
Cumulative size distribution for outlet Andersen run
(run no. M5B-0-R1) at Du Pont Boiler No. 3 in Parkers-
burg, WV (December 17, 1980).
-------
TABLE 8a. SUMMARY OF METHOD 9 PLUME OPACITY OBSERVATIONS AT
DU PONT BOILER NO. 3 IN PARKERSBURG, WV
(DECEMBER 16, 1980)
Date: 12-16-80
Type of Discharge: Stack
Height of Point of Discharge:
Wind Direction: S.W.
Color of Plume:
Light gray to
dark gray
Type of Plant: Industrial boiler
Location of Discharge: Mechanical collector exhaust
180 ft Description of Sky: Overcast
Wind Velocity: 3-5 mph
Above mouth of stack
Location of Observation
Duration of Observation
Observer Name: Thomas Malone
Distance from Observer to Discharge Point: 900 ft
Direction of Observer from Discharge Point: S.E.
Height of Observation Point: Ground level
Sampling Time:
360 min total
a
1 11:40-14:32
2 17:13-20:00
Summary of average opacity
Set
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Time
Start
11:10
11:16
11:22
11:28
11:34
11:40
11:46
11:52
11:58
12:04
12:10
12:16
12:22
12:28
12:34
12:40
12:46
12:52
12:58
13:04
13:10
13:16
13:22
13:28
13:34 ,
13:40
13:46
13:52
13:58
14:04
End
11:15
11:21
11:27
11:33
11:39
11:45
11:51
11:57
. 12:03
12:09
12:15
12:21
12:27
12:33
12:39
12:45
12:51
12:57
13:03
13:09
13:15
13:21
13:27
13:33
12:39
13:45
13:51
13:57
14:03
14:09
Opacity
Sum
325
270.
K
315
375
425
450
540
440
495
450
460
350.
h
340b
240D
_c
K
170°
4
-------
TABLE 8a (continued)
I
25
a.
30
20 -
10-
Summary of average opacity
Set
number
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Test ended
Average, all
Time
Start
14:10
14:16
14:22
14:28
14:34
14:40
14:46
14:52
14:58
15:04
15:10
15:16
15:22
15:28
15:34
15:40
15:46
15:52
15:58
16:04
16:10
16:16
16:22
16:28
16:34
16:40
16:46
16:52
16:58
17:04
sets except
End
14:15
14:21
14:27
14:33
14:39
14:45
14:51
14:57
15:03
15:09
15:15
15:21
15:27
15:33
15:39
15:45
15:51
15:57
16:03
16:09
16:15
16:21
16:27
16:33
16:39
16:45
16:51
16:57
17:03
17:09
15 and 19
Opacity
Sum
390
405
410
495.
h
90b
260b
390
430
220
360
445.
h
420°
465
405
430
420
420
390
345
485
455
340
385
480
450
500
535
525
540
495
17.1
Average
16.3
16.8
17.1
20.6.
n
6-4b
16. 3b
16.3
17.9
9.2
15.0
18.5.
n
18.3
19.4
16.9
17.9
17.5
17.5
16.3
14.4
20.2
19.0
14.2
17.5
20.0
18.8
20.8
22.3
21.9
22.5
20.6
i/*VAvr1orw-
0 1
2
3
4
5 6
TIME, hours
18
-------
TABLE 8b. SUMMARY OF PLUME OPACITY OBSERVATIONS AT DU PONT BOILER
NO. 3 IN PARKERSBURG, WV (DECEMBER 17, 1980)
Date: 12-17-80
Type of Discharge: Stack
Height of Point of Discharge
Wind Direction: S.W.
Color of Plume
Type of Plant: Industrial boiler
Location of Discharge: Mechanical collector exhaust
180 ft Description of Sky: Partly cloudy
Wind Velocity. 2-4 mph
Location of Observation: Above mouth of stack
Observer Name: Thomas Halone
Dark gray, changing
to light brown
at 12:30
Duration of Observation:
Distance from Observer to Discharge Point: 900 ft
Direction of Observer from Discharge Point: S.E.
Height of Observation Point: Ground level
420 min total
a
Sampling Time:
10:18-13:01
15:15-19:10
Summary of average opacity
Set
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Time
Start
9:50
9:56
10:02
10:08
10:14
10:20
10:26
10:32
10:38
10:44
10:50
10:56
11:02
11:08
11:14
11:20
11:26
11:32
11:38
11:44
11:50
11:56
12:02
12:08
12:14
12:20
12:26
12:32
12:38
12:44
End
9:55
10:01
10:07
10:13
10:19
10:25
10:31
10:37
10:43
10:49
10:55
11:01
11:07
11:13
11:19
11:25
11:31
11:37
11:43
11:49
11:55
12:01
12.07
12:13
12:19
12:25
12:31
12:37
12:43
12:49
Opacity
Sum
370
320
285
410
335
155
305
245
385
455
305
340
445
41 5b
440
665
765.
h
390
640.
K
545D
645.
Jj
595K
410b
445b
370b
475
630
410.
455b
Average
15.4
13.3
11.9
17.1
14.0
6.5
12.7
10.2
16.0
19.0
12.7
14.2
18.5
17.8
22.0
27.7
31.9.
K
32. 5D
26.7.
h
23. 7D
28.0
20 -5K
K
25 -9h
22. 8^
21. ij
16.1b
19.8
26.3
17.1.
19. 8b
(continued)
19
-------
TABLE 8b (continued)
wind Direction: N.E.
Wind Velocity: 3-5 mph
Color of Plume: Light brown
Summary of average opacity
Set
number
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Time
Start
12
12
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
:50
:56
:02
:08
:14
:20
:26
:32
:38
:44
:50
:56
:02
:08
:14
:20
:26
:32
:38
:44
:50
:56
:02
:08
:14
:20
:26
:32
:38
:44
End
12
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
15
15
15
15
15
15
15
15
15
:55
:01
:07
:13
:19
:25
:31
:37
:43
:49
:55
:01
:07
:13
:19
:25
:31
:37
:43
:49
:55
:01
:07
:13
:19
:25
:31
:37
:43
:49
Opacity
Sum
585
585
700
660
625
480
570
565
505
435
550
555
410
385
515
670.
n
705
635.
K
410°
540
260d
675
640
575
530
550
600
525
520
445
Average
24
24
29
27
26
20
23
23
21
18
22
23
17
16
21
27
30
26
19
22
20
28
24
24
22
22
25
21
21
18
.4
.4
.2
.5
.0
.0
.8
.5
.0
.1
.9
.1
.1
.0
.5
.9.
K
.7°
.5.
K
.5°
-5.
.od
.1
.7
.0
.1
.9
.0
.9
.7
.5
(continued)
20
-------
TABLE 8b (continued)
Description of Sky: Clearing
Summary of average opacity
Set
number
61
62
63
64
65
66
67
68
69
70
Test ended
Time
Start
15:50
15:56
16:02
16:08
16:14
16:20
16:26
16:32
16:38
16:44
End
15:55
16:01
16:07
16:13
16:19
16:25
16:31
16:37
16:43
16:49
Opacity
Sum
430
490
510
620
530
595
700
690
830
845
Average
17.9
20.4
21.3
25.8
22.1
24.8
29.2
28.8
34.6
35.2
Average, all sets: 21.9
Total observation time includes periods when no opacity readings could be obtained
because of transient conditions.
Steam from an adjacent stack prevented opacity readings from being taken during
part of this observation period.
No opacity readings could be taken during this observation period because
of steam interference.
Observations were interrupted for 2 min and 45 s by plant security personnel
seeking identification.
3 a
TIME, hours
21
-------
TABLE 9. SUMMARY OF COAL ANALYSIS AT DU PONT BOILER
NO. 3 IN PARKERSBURG, WV (DECEMBER 16-17, 1980)
Quality assurance sample*
Andersen Analysis Analysis
runs at with with
Parameter Run 1 Run 2 Run 3 Run 4 inlet Parkersburg Parma Difference
Total moisture. %
Heating value, J/q
Dry basis
As received
Ash. %
Dry basis
As received
Sulfur, %
Dry basis
As received
Carbon, %
Dry basis
AS received
Hydrogen, %
Dry basis
As received
Oxygen. %
Dry basis
As received
Nitrogen, %
Dry basis
As received
Sieve sizing sample weight, g
1.35
3.64
3.92
3.10
31.800 32,800 32.100 31,800
31,400 31,600 30,700 30,900
7.60
7.50
2.69
2.66
77.70
76.65
4.99
4.93
5.60
5.53
1.41
1.39
7.72
7.44
2.70
2.60
77.43
76.61
5.41
5.21
5.19
5.00
1.55
1.49
7.58
7.29
2.72
2.61
77.88
74.82
5.73
5.50
4.70
4.52
1.39
1.34
8.68
8.41
2.86
2.77
75.40
73.07
5.12
4.97
6.63
6.43
1.30
1.26
1,307.2 1,250.9 1,283.4 1,218.8
3.28
32,100
30,900
7.77
7.52
2.88
2.79
77.25
74.72
5.19
5.02
5.41
5.23
1.33
1.29
359.9
4.45
31,800
30,200
8.05
(7.90)*
7-72
(7.55)
2.6?
(2.60)*
2'5i
(2.48)*
75.99
72.61
5.25
5.02
6.82
6.51
1.27
1.22
414.5
3.54
33,500
32,300
3 9i
(4.10)*
3.78
(3.96)*
(1-29)"
l'3§
(1.24)*
80.9
76.1
5.9
5.5
8.6
8.0
1.14
1.07
585.5
0.91
-1,600
-2,100
4.13
3.92
1.27
1.20
-4.91
-3.49
-0.65
-0.48
-1.78
-1.49
0.13
0.05
171
Percent by weight collected on sieve sizes
25.4 mm sieve (1.0 in.)
1)4 sieve, 4.75 ran (0.187 in.)
t»16 sieve, 1.18 mm (0.046 in.)
#30 sieve. 0.60 mm (0.024 in.)
»200 sieve, 0.075 mm (0.003 in.)
Passing through all sieves
Cumulative percent by weight less than sieve sizes
25.4 mm (1 in. sieve)
«4 sieve. 4.74 mm (0.187 in.)
K16 sieve, 1.18 mm (0.046 in.)
»30 sieve. 0.60 mn (0.024 in.)
11200 sieve, 0.075 mn (0.003 in.)
0
92.20
4.76
0.76
1.81
0.47
100
7.8
3.04
2.28
0.47
80
12
2
3
0
19
7
4
0
0
.38
.87
.59
.93
.73
100
.62
.25
.66
.73
1
88
4
2
2.
0
98
.10
" 5
3.
o.
.09
.62
.78
.33
.70
.48
.91
.29
.96
.18
.48
76
15
2
3
0
13
7
4
0
0
.99
.96
.78
.37
.90
100
.01
.05
.27
.90
0
60.30
27.18
7.93
4.46
0.13
100
39.7
12.52
4.59
0.13
21.84
57.38
10.29
4.03
6.46
0.0
78.16
20.78
10.49
6.46
0.0
11.90
74.75
6.93
1.79
3.76
0.75
88.1
13.23
b.3
4.51
0.75
9.94
17.37
3.36
2.24
2.7
-0.75
-9.94
7.55
4.19
I .95
-0.?5
The results in parentheses are from a second analysis of the homogenized, pulverized coal samples previously analyzed after
sample preparation. The numbers not in parentheses are from the analysis of different halves of a raw coal sample that was split
with a riffle sampler.
To convert to Btu/lb, multiply J/g by 4.3 x 10 l.
22
-------
differences in sulfur content were between 0.02% and 0.06% (dry)
and the differences in ash content were between -0.18% (dry) and
0.15% (dry). This suggests that the poor reproducibility of the
coal sulfur and ash content results in Table 9 may be caused by
real differences in the composition of coal from one piece to
another and the small amounts of sample that were split in half,
prepared, and analyzed.
Appendix F-4 contains a summary of available information on the
precision and accuracy of the sulfate and sulfuric acid analysis.
Four quality control samples of sulfuric acid were prepared from
0.02 N sulfuric acid and analyzed along with the Parkersburg
stack samples. The measured amounts of sulfuric acid were betwen
81% and 120% of the reported "true" values. The results of the
analysis of four EPA sulfate quality control filter strip samples
showed the measured values to be between 101% and 118% of the
"true" amounts of sulfate. The possibility exists of differences
in the physical and chemical composition of the stack samples
(compared to those of the quality control samples) changing the
accuracy of analysis of the stack samples from that reported for
the quality control samples.
23
-------
SECTION 3
PROCESS DESCRIPTION AND OPERATION
The powerhouse at the Du Pont Washington Works plant consists of
six steam boilers, all coal-fired, spreader stoker fed. The total
combined steam generating capacity is about 284,000 kg/hr
(625,000 Ib/hr). All six boilers produce nominal 186,000 kg/m2
gage (265 psig) saturated steam and are interconnected by the
common plant steam supply header. Particulate matter flue gas
emissions from boilers No. 1 and No. 3 are controlled by dual
multi-tube cyclone dust collectors; the other boilers are con-
trolled by single multi-tube cyclones followed by baghouse filters
Boiler No. 3, with a nominal steam production capacity of
34,000 kg/hr (75,000 Ib/hr), was the only boiler sampled and
monitored during the test. Boiler No. 3 shares a smokestack with
boiler No. 4. Boiler No. 4 was shutdown during the test to
permit visible emission stack readings for boiler No. 3.
COAL HANDLING SYSTEM
Bituminous coal with a sulfur content ranging from 2 to 3 percent
is burned in all six boilers. Coal is delivered to the plant as
6 mm to 38 mm (1/4" to I 1/2") chunks in railroad hopper cars.
Once per 8-hour shift, a belt conveyor transfers coal from stor-
age piles or directly from the bottom of a railroad car to
crushers. The belt conveyor has an integrated weighing mechanism
to determine the total amount of coal transferred to the boilers.
The crushed coal, ranging in size from 6 mm to 19 mm (1/4" to
3/4", with 15% to 20% less than 6 mm), is then distributed by
bucket elevators and another belt conveyor to individual coal
25
-------
feed bunkers for each boiler. The bunkers are instrumented with
high and low level indicators, but no weighing device instrumen-
tation is used to determine the actual coal feed rate to the
boiler.
BOILER OPERATION
Boiler No. 3 is a watertube design, manufactured in 1957 by
Badenhausen Co. (a division of Riley Stoker Corp.). The original
design did not include an economizer which was added several
years later. The total design heat input with the economizer is
rated at about 10 billion Joules/hour (93 million Btu/hr) based
on a nominal boiler efficiency of 84 percent for generating
34,000 kg/hr (75,000 Ib/hr) of steam.
Operation of boiler No. 3 requires continuous coal and combustion
air feeds to the furnace, as well as continuous bottom ash removal
and venting of the flue gas. A continuous boiler feed water
(BFW) flow and boiler blowdown are also required to satisfy the
constant plant steam demand. A schematic flow diagram of boiler
No. 3 is presented in Figure 4.
Coal feed to the boiler is accomplished through a combination of
gravity and mechanical rotating feeders. Crushed coal moves its
way down from the bunker and is directed to three Riley spreader
stoker units located on the furnace front. These rotating feeders
distribute the coal by flinging the particles up over the flames
to the back of the furnace fire. The coal fire is supported by a
8 to 13 cm (3" to 5") layer of ash on a continuously travelling
grate. The grate slowly moves the ash bed back up to the furnace
front where the particles fall off and accumulate in a collector
hopper under the boiler. The hopper has about a 16-hour capacity
and is routinely unloaded every eight hours.
26
-------
Preheater
Coal
Feed
Bunker
Economizer
to
Coal Feeder
Spreader
Stoker
265 PSIG
Steam Export
To
Deaerator
Induced
Draft
Fan
(
Stack
No. 2
<«&> 1 ^
uM
M "5
Two-Stage
Mechanical
Collector
To
-^ t Overf i re Ai r
>*-*>. Blower
Forced Draft
Combustion Air
Fan
Steam
Turbine
Exhaust
Ash
Collection System
To
Bottom Ash
Collection System
I
Vent
J_
Flash
Tank
Boiler
Feed -
Hater
•a
Deaerator
Condensate
Recycle From
Production
Areas
1
....J_..._
unijunn
Slowdown
Discharge to
River
Heat Recovery
Exchanger
Figure 4. Process flow schematic of Du Pont Washington Works Boiler No. 3.
-------
The total combustion air fed to the boiler furnace consists of a
main combustion air supply and a mixing air supply. The forced
draft fan, with a nominal capacity rating of about 880 std m3/min
(31,000 std ft3/min), feeds ambient air up through the bottom of
the furnace fire. A much smaller overfire air blower injects
ambient air over the top of the furnace fire to provide adequate
turbulence for complete combustion and heat transfer.
The BFW supply consists of condensate recycled from the plant
production areas combined with softened make-up well water. The
cold well water is first pumped through a boiler blowdown heat
exchanger to recover some heat from the discharged blowdown. The
warmed well water is then combined with the recycled condensate
and directed through a deaerator and a steam preheater. The
preheater is designed to heat the BFW up to about 135°C (275°F)
by controlling the steam flow. The preheated BFW is then directed
through an economizer installed in the flue gas duct from the
boiler furnace. The economizer is designed to increase the
boiler thermal efficiency by recovering some of the sensible heat
in the flue gas before it is discharged to the atmosphere. The
addition of the economizer to the original boiler design
increased the boiler thermal efficiency rating from 77 to about
84 percent. From the economizer, the hot feed water is directed
to the boiler steam drum.
The steam produced is controlled at 19 kg/cm2 (265 psig). The
saturated steam will have a temperature of about 210°C (410°F).
The boiler is not designed to produce any superheated steam. The
nominal design steam production rate is about 34,000 kg/hr
(75,000 Ib/hr).
The accumulation of salts and other foreign material in the
boiler internals is minimized by a dual blowdown system. The
main mechanism is a continuous water blowdown from the steam
drum. The blowdown rate is continuously controlled by an on-line
28
-------
conductivity analyzer. According to plant personnel, the contin-
uous blowdown rate is normally equal to about 8 to 12 percent of
the BFW flow rate. The secondary blowdown mechanism involves the
opening, for about 15 seconds, once per day, of a manual valve in
the water drain line from the mud drum. Both blowdown streams
are directed through a flash tank and a heat recovery exchanger
before being discharged to the river.
FLUE GAS/AIR POLLUTION CONTROL SYSTEM
Flue gas from the boiler is pulled through an economizer and a
two-stage mechanical dust collector by the induced draft fan.
The induced draft fan and the forced draft fan share a common
steam turbine drive. The induced draft fan, with a nominal
capacity rating of about 1,700 m3/min (60,000 ft3/min) at 366°C
(690°F), discharges the flue gas through stack No. 2. Potential
flue gas particulate emissions are reduced by both the economizer
and the mechanical dust collector. There are no other air pollu-
tion controls for the flue gas from boiler No. 3.
The economizer is a specially designed multi-tube heat recovery
exhanger. The flue gas passes down over one half of the tube
bundle and then up through the second half. BFW flowing through
the tubes cools the flue gas which causes some particulate matter
to drop out. This removed particulate matter, or fly ash,, is
collected in a hopper at the bottom of the economizer. The
hopper has about a 16-hour capacity and is routinely unloaded
every eight hours. Under normal operations, the boiler exit flue
gas temperature ranges from about 290° to 320°C (560° to 615°F).
The economizer exit flue gas temperature ranges from about 175°
to 185°C (345° to 365°F).
Most of the potential particulate emissions are controlled by the
multi-tube cyclone mechanical dust collector. The collector,
manufactured by Universal Oil Products Co., consists of two stages
29
-------
Each stage contains 200, 15 cm (6") diameter cyclonic elements.
The collector has no moving parts and uses no scrubbing media.
Separation of the particles from the flue gas is accomplished by
centrifugal force, generated by directing the gas stream through
the static cyclonic elements. The removed particulate matter is
collected in hoppers at the bottom of each collector stage in a
similar fashion as for the economizer. The hoppers to the col-
lectors have about 16-hour capacities and are routinely unloaded
every eight hours. Previous emission tests were conducted at
full boiler capacity in 1974 for the State of West Virginia. The
results showed that the first stage had a particulate removal
efficiency of about 97 percent, with the second stage efficiency
at about 52 percent, for an overall efficiency exceeding 98 per-
cent for both stages. At full boiler load, the total pressure
drop across both stages is normally about 12 cm to 15 cm (5" to
6") of water.
A Lear Siegler RM4 transmissometer is installed in the flue gas
stream just before it enters the stack. This device provides an
on-line, continuous relative indication of the particulate removal
efficiency by monitoring the flue gas opacity.
ASH HANDLING SYSTEM
The ash handling facilities consists of two independent pneumatic
conveying systems. One system handles the bottom ash from the
boiler furnace. The second system handles the fly ash removed
from the flue gas. A cross-over is installed between the two
facilities in case one system is required to handle both services.
A separate ash silo is provided for each facility. The ash is
disposed of by landfilling.
The bottom ash handling system consists of a batch, manual step
combined with a pneumatic, vacuum conveying system. About every
eight hours the bottom ash is manually raked out of the collection
30
-------
hopper and down into a large funnel connected to the pneumatic
transfer line. The motive force for the ash transfer is provided
by a vacuum pump. The vacuum pump pulls the air/ash mixture
through the transfer line to a cyclonic separator. The separated
ash falls down into a silo. The air from the separator is pulled
through a fabric filter before being discharged to the atmosphere.
The fly ash handling system consists of facilities identical to
that for the bottom ash except that fly ash transfer is com-
pletely automatic, with no manual labor required. The vacuum
line for transferring the air/ash mixture is directly connected
to the bottom of the economizer and mechanical dust collector
hoppers through remote operated valves. About every eight hours
a pre-programmed sequencer automatically signals the valves to
open for transferring the fly ash.
OPERATING CONDITIONS
The boiler operated in fairly smooth, normal fashion during the
four, three-hour test runs conducted during December 16 and 17.
The first three runs were conducted near the full production rate
of 35,000 kg/hr (77,000 Ibs/hr) steam. The fourth run was con-
ducted to observe the effects on emission rates at reduced boiler
load. This latter run was at 75 to 80 percent rate, which was
the minimum achievable boiler loading allowed by the plant steam
demand. To prevent possible boiler upsets, the powerhouse opera-
tors agreed not to remove bottom ash, remove fly ash, or open any
doors on the furnace front during the test. Soot blowing is not
used on this boiler. All of these necessary procedures vrere
completed before testing was started.
Seventeen process parameters were monitored during the emission
tests. The raw data sheets are presented in Appendix D. These
parameters are presented in Table 10 and discussed below.
31
-------
TABLE 10. SUMMARY OF PROCESS CONDITIONS DURING TESTING
AT DU PONT BOILER NO. 3 IN PARKERSBURG, WV
(DECEMBER 16-17, 1980)
Parameter
range/average 1
Monitored parameters
Boiler feed water
-Flow. 1,000 Ib/hr 74 - 84
79.4
-Economizer inlet temp., °F 280
-Slowdown cond. , pmho's 3,800
Steam production
-Flow, 1,000 Ibs/hr 71 - 78
76.7
-Pressure, psig 265
Forced draft (FD) fan,
in. water pressure +0.6
Overfire air blower, in. water
pressure +18
Furnace pressure, in. water
(draft) -0.07
Flue gas
-Furnace outlet pressure
in. water (vacuum) -1.6
-Furnace outlet temp., °F 610
-Oxygen, % 5.0 - 6.4
5.8
-Economizer outlet pressure
in. water (vacuum) -4.5
-Economizer exit temp. , °F 360
-Mech. dust collector pres-
sure drop, in. water 6.5
-Stack opacity meter, % 11-14
12
Induced draft (ID), fan
-Driver speed, RPM (4X) 5,100
-Diff. pressure, in. water 7.7
Derived parameters
Test
2
78 - 84
80.8
280
3,700
73 - 80
78.5
265
+0.6
+ 18
-0.07
-1.6
610
4.8 - 6.0
5.4
-4.5
360
6.6
11 - 13
12
5,100
7.7
run
3 4
76 - 81 60 - 64
79.1 62.6
280 288
3,400 3,000
76 - 78 58 - 62
76.9 59.6
265 265
+0.5 +0.3
+19 +18
-0.08 -0.07
-1.6 -0.9
615 560
5.0 - 5.6 6.2 - 7.2
5.3 6.7
-4.5 -2.8
365 345
6.6 4.0
12 7
5,100 3,800
7.9 4.8
Boiler blowdown rate, Ib/hr 2,700 2,300 2,200 3,000
Flue gas flow rate by ID fan
test curve, dscfm 28,470 28,260 27,450 16,720
Boiler heat input of coal,
million Btu/hr 89.6 90.7 89.0 66.3
32
-------
1. BFW Flow Rate- The boiler feed water flow rate was
monitored by two instruments on the boiler control
panel. The instantaneous rate was monitored on a flow
recorder. The average flow rate over each test run was
obtained from a totalizer. Periodic rate checks between
the totalizer reading differences and the flow recorder
readings showed that the totalizer read 900-1800 kg/hr
(2,000 to 4,000 Ibs/hr) higher flow. The plant instru-
ment maintenance personnel suggested that the totalizer
was more accurate than the flow recorder because the
flow recorder has a square root extractor to linearize
the instrument signal. With this assumption, the flow
rate range and average during each test run is presented
in Table 10. The BFW flow is directly affected by
fluctuations in both the steam flow and blowdown flow.
BFW flow rate fluctuations are, therefore considered
normal operation. The average flow rate during run
No. 4, the lower rate run, was about 78 percent of the
first three runs.
2. BFW Temperature - The boiler feed water temperature was
continuously monitored between the steam preheater and
the economizer by a local temperature indicating
controller. The temperature remained very constant
during all four test runs. There was no temperature
monitor available between the economizer outlet and the
boiler steam drum.
3. Boiler Blowdown Conductivity - The boiler blowdown was
continuously sampled and monitored by an on-line
analyzer. This parameter indicates whether there is
sufficient blowdown - high conductivity would indicate
accumulation of salts, etc. The conductivity varied
somewhat between test runs, but was very constant
during each run.
33
-------
A monitoring device for the boiler blowdown rate was
not available. However, this flow rate can be derived
by calculation as the difference between the BFW and
the steam flow rates. The blowdown rate was expected
to be about 2,700 to 3,600 kg/hr (6,000 to 8,000 Ibs/hr)
Plant personnel suspected that the flow meters are not
accurate enough to calculate the actual blowdown rate.
•
4. Steam Production Rate - The steam flow rate was moni-
tored by three instruments on the boiler control panel.
The instantaneous rate was monitored on a flow recorder
and shown on a CRT output (updated about every five
minutes) for the plant computer monitor. The average
flow rate over each test run was obtained from a
totalizer. Periodic rate checks between the totalizer
reading differences and the flow recorder readings
showed that the totalizer read 1,800 to 2,700 kg/hr
(4,000 to 6,000 Ibs/hr) higher flow. Flow indication
by the computer agreed very well with the totalizer.
Thus, it appears that the flow recorder is not as
accurate for the same reasons as explained for the BFW
flow recorder. With this assumption, the steam pro-
duction rate during each test run is presented in
Table 10. A slight fluctuation in the steam flow is
considered normal operation. The average flow rate
during run No. 4 was about 77 percent of the first
three runs.
5. Steam Pressure - The steam pressure was monitored by a
local pressure gauge and by one inside the control
room. The plant personnel considered the local gauge
to be more accurate. The steam pressure was very
steady during each test run.
34
-------
6. Forced Draft Fan - The discharge pressure of the forced
draft fan was monitored by a gauge on the control
panel. This pressure remained very steady during each
test run, indicating steady air flow to the furnace.
7. Overfire Air Blower - The discharge pressure of the
overfire air blower was monitored by a gauge on the
control panel. This pressure remained very steady
during each test run, indicating steady air flow to the
furnace.
8. Furnace Pressure - The furnace pressure was monitored
by a gauge on the control panel. The furnace is nor-
mally under a slight negative pressure or draft. The
furnace draft remained fairly steady - slight fluctua-
tions is considered normal operation.
9. Furnace Outlet Pressure - The flue gas pressure exiting
the boiler furnace was monitored by a gauge on "the
control panel. This pressure is normally lower, or
more negative, than the furnace pressure. This pres-
sure remained very steady during each test run.
10. Furnace Outlet Temperature - The flue gas temperature
exiting the boiler furnace was monitored by a recorder
on the control panel. This temperature remained almost
constant during each test run.
11. Flue Gas Oxygen - The percent oxygen in the flue gas
was monitored by a recorder on the control panel. The
flue gas at the furnace outlet was continuously sampled
for the oxygen content by an on-line analyzer. The
oxygen content fluctuated quite a bit as shown in
Table 10. However, these fluctuations are considered
normal operations, so long as the oxygen content remains
above five percent.
\
35
-------
12. Economizer Outlet Pressure - The flue gas pressure out
of the economizer was monitored by a gauge on the
control panel. This pressure is normally lower, or
more negative, than the furnace outlet pressure. This
pressure remained very steady during each test run.
13. Economizer Exit Temperature - The flue gas temperature
exiting the economizer was monitored by a recorder on
the control panel. This temperature remained very
steady during each test run.
14. Mechanical Dust Collector Pressure Drop - The flue gas
pressure drop through both stages of the multi-cyclone
collector was monitored by a recorder on the control
panel. This pressure drop will vary with the flue gas
flow rate, or boiler loading, the particle size distri-
bution and concentration, and any flow restrictions due
to build-up of particles on the cyclone internals.
This pressure drop, fluctuated only slightly during each
test run. The pressure drop was much lower during run
No. 4, the lower rate run, than during the first three
runs.
The day before testing started, the pressure drop was
abnormally high at over 8 inches of water. Plant
personnel mentioned that the high pressure drop might
be caused by mud mixed in with the coal resulting in
excessive amounts of particulate matter in the flue
gas. The coal feed was changed and the pressure drop
returned to the normal range before testing started.
15. Stack Opacity - The percent opacity in the flue gas was
monitored by a recorder on the control panel. The
opacity of the flue gas after the mechanical collector
36
-------
was continuously monitored by an in-stack opacity
meter. The meter was set-up with a periodic air purge
to minimize particulate accumulation on the sensor
element. Small fluctuations in the stack opacity are
considered normal operation. The opacity was much
lower during run No. 4, indicating increased mechanical
collector efficiency, despite the lower flue gas flow
rate.
16, 17. Induced Draft Fan - The steam turbine driver speed was
monitored by a portable resonant-reed tachometer. The
actual fan speed is reduced to one-fourth the driver
speed by a gear reducer. The driver speed remained
almost constant during each test run, indicating steady
flue gas flow. The fan speed was decreased during the
low rate run No. 4. This same steam turbine drive and
gear reducer also serves the forced draft fan.
The differential pressure across the induced draft fan
was monitored by a portable manometer connected to the
suction and discharge ductwork. This differential
pressure remained almost constant during each test run.
The differential pressure was much lower during run
No. 4 because of the lower fan speed.
The boiler performance test results are in excellent
agreement with the 84 percent Du Pont nominal effi-
ciency rating. For more details, see Appendix D.
37
-------
SECTION 4
LOCATION OF SAMPLING POINTS
Air pollution emissions at two general locations along the exhaust
path of Boiler No. 3 were tested: a site between the economizer
and the first-stage mechanical collector (the control system in-
let), and a site after the second stage of the mechanical collec-
tor and an exhaust fan, on a duct leading to a smokestack (the
control system exhaust). The general locations of these sampling
sites are shown in Figure 5. The duct at the inlet is about 4.4
meters (14 ft, 4 in.) deep, and it had to be sampled from its
north side and its south side. Figure 6 shows a top view of the
inlet sampling sites. In addition, the coal being fed into the
boiler was sampled.
CONTROL SYSTEM INLET (NORTH SIDE)
Figure 7 contains a schematic diagram of the sampling site on the
north side of the inlet to the first-stage mechanical collector
and a photograph of the site before the grating was extended for
testing. Four 10 cm (4 in.) ID sampling ports are located on a
51 cm (20 in.) long length of straight ductwork at the end of a
180° turn at this site. The duct has inside dimensions of
1.19 meters (47 in.) by 4.37 meters (14 ft, 4 in.). Two straight-
ening vanes are located within the elbows on both sides of the
sampling ports (see Figure 8). The vanes between the boiler and
the ports were within 5 cm (2 in.) of the two middle sampling
ports. The 180° turn is located immediately above the boiler
economizer and the first-stage mechanical collector. The site
does not meet the criteria for an acceptable measurement site for
39
-------
Figure 5. Schematic diagram of Du Pont Boiler No. 3
associated equipment in Parkersburg,
and
-------
DOOR
o
Q
m
°0
°0
o
PIPES
O
LADDER
/
NORTH INLET PORTS
FLOW
4.37 m
(14ft-4in.)
c~
SOUTH INLET PORTS
LADDER
Figure 6.
Top view of inlet sampling area at Du Pont
Boiler No. 3 in Parkersburg, WV.
41
-------
SIDE VIEW
137 cm 0. D.
(54 in.)
I
23cm (9in.)
30cm (12in.)
30cm (12in.)
30cm (12in.)
18cm (7 in.)
--O
FIRST STAGE
COLLECTOR
119 cm I. D.
(47 in.)
51cm
ECONOMIZER
Figure 7. North side inlet sampling site at Du Pont
Parkersburg, WV Boiler No. 3.
-------
,-'&&w$%i.
. rw-1* ..•*• '*«'•-/ .•? ••I* • v ..-'•• . •.:
+r".ir ' • \ ' 'f ' +**^Alt' *' *^*' «•"• .".
tS'VT^ \ '/:• -Si,-" -
Figure 8. Inside view of inlet duct at Du Pont
Boiler No. 3 in Parkersburg, WV.
43
-------
particulate traverses under EPA Method 1 and Method 5 procedures,
but no better sites were available without modifying the ductwork.
Twelve points along each of four diameters of the duct were
sampled through the available ports. Half of the points were
sampled from the north side and half from the south side.
An Andersen particle size sample was taken through the second
port from the top of the duct, 91 cm (3 ft) from the inside of
the north wall of the duct. This location had a flue gas veloc-
ity approximately equal to the arithmetic mean of the velocities
at the 24 traverse points in the north half of the duct.
CONTROL SYSTEM INLET (SOUTH SIDE)
The ductwork and sampling port configuration on the south side of
the duct are the same as on the north side. Figure 9 contains
schematic diagrams and a photograph of this side of the duct
before the grating was extended for testing. The sampling points
here suffer from the same problems encountered on the north side
(inadequate lengths of straight ductwork and straightening vanes
located within 5 cm (2 in.) of the two middle sampling ports).
An Andersen particle size distribution sample was taken 91 cm
(3 ft) from the south inside wall of the duct through the third
port from the top. The flue gas velocity was approximately equal
to the average velocity at the 24 traverse points in the south
half of the duct at this location.
CONTROL SYSTEM EXHAUST
Figure 10 contains a photograph of the duct between the exhaust
fan and the stack and a drawing showing some of the relevant
dimensions of the ductwork. Six 10 cm (4 in.) ID sampling ports
are located on a 137 cm (54 in.) long length of rectangular duct
with inside dimensions of 102 cm (40 in.) by 131 cm (51.6 in.).
The ports are 81 cm (32 in.) downstream from a slight flaring of
44
-------
80IIER
SOOTH SIK OF INLET
INLET DUCT *. 57m (15 ft) DEEP
L. 127cm j
~
(50 In.) I
._ 23cm (9In.I
|r 30cm (12 In. I
;T 30_cm (12 In. I
:: 3o7m (12 In. I
II cm (7 In. I
Or CLONE II
ECONOMIZER
Figure 9. South side inlet sampling site at Du Pont
Parkersburg, WV Boiler No. 3.
-------
10cm (4 in.)
•/ P!PE COUPLINGS
® ,D
FLOW
STACK NO. 2
GROUND LEVEL-
Figure 10.
Exhaust sampling site at Du Pont
Parkersburg, WV Boiler No. 3.
-------
the ductwork to form a transition to the fan and 56 cm (22 in.)
upstream from a flaring of the ductwork to connect with the
stack. The stack is used by another boiler also. The site does
not meet the criteria for an acceptable measurement site for par-
ticulate traverses under EPA Method 1 and Method 5 procedures.
The State of West Virginia accepted a compliance test using data
taken at this site a few years ago, and a decision was made by
EPA to measure emissions at this location using eight treiverse
points along each of six diameters of the duct through the avail-
able sampling ports. A special dual-probe stack sampling train
was used to permit sampling by Methods 5 and 5B through the same
ports at the same time, with the nozzles within 4.4 cm (1.75 in.)
of each other.
The Andersen particle size sample was taken through Port C (shown
in Figure 10) at a location having a velocity approximately equal
to the arithmetic mean of the velocities measured at the 48
Method 5 and 5B traverse points. This location is 107 cm (42 in.)
from the inside wall of the duct.
COAL SAMPLING SITE
Grab samples of coal were taken with a shovel through three sam-
pling ports at the bottom of a coal feed chute leading directly
to the boiler coal feeder overthrow rotor. The coal samples were
taken before, during, and after every Method 5 and 5B emission
test.
47
-------
SECTION 5
DESCRIPTION OF SAMPLING TRAINS
PARTICULATE SAMPLING TRAIN
A special dual-probe stack test system was used for stack testing
by Methods 5 and 5B simultaneously through the same individual
stack ports. Figure 11 is a schematic of the dual-probe system.
Heated glass-lined probes and Reeve Angel Type 934 AH filters
were used for the Method 5 and 5B testing. The filter tempera-
tures were monitored using thermocouples installed on the back
half of the filter holders. The nozzle centers of the Method 5
and SB trains were approximately 4.4 cm (1.75 cm) apart in this
system. A single pitot tube between the Method 5 and 5B sampling
probes was used to measure the flue velocity for both trains.
Method 3 integrated bag samples were taken from the exhaust of
the Method 5 train. Method 5 and 5B sample volumes of at least
1.2 dry m3 (44 ft3) were taken. When Methods 5 and 5B were
used, Methods 1 through 4 were also used if possible, as in
typical compliance monitoring. Typical Method 5B trains, heated
to 320°F ± 25°F, were used at the inlet locations.
PARTICULATE SIZE SAMPLING TRAIN
Sampling for particle size was performed using an Andersen cas-
cade impactor with seven stages and a back-up filter. Figure 12
is a schematic drawing of the sampling train. It consisted of
the following equipment, listed in order of the air flow: a
probe tip with a diameter between 9.5 mm and 9.8 mm (3/8 in. and
3/16 in.) depending on flue gas velocity; a 3.0 m (10 ft) glass-
lined probe [or a 1.2 m (4 ft) stainless steel probe]; a condenser;
49
-------
FUXlItt
(MPINGERS
FILTER
THERMOCOUPLE
S-TYPE
PITOTTUBE HEATED PROBE
VACUUM LINE
DRY GAS AIR-TIGHT
METER PUMP
Figure 11. Dual-probe sampling train.
50
-------
Thermometer
r\
Cascade
Impaclor
Nozzle
t
Reverse-Type
Pilot Tube
Vacuum
Tubing
Slack Wall
Probe
J
1 ;•
;e Bath—*-
Water -
^
T
U-
i i . > <
r (
_/
<$
.
~
u
n
j'
'•,-.
i
^
if.
>.
«-•-
^}
**
-*-
v y
Silica
Gel
Pilot Manomeler^Thermome,ers BVPass
Orifice
\7
Impingers
0T\
y fl Vacuum Gauge
C*3=q| —
t
Check Valve
Vacuum
Line
Main Valve
Air Tight Pump
Dry Gas Meier
Figure 12. Particle size distribution sampling apparatus - Andersen 2000, Inc.
-------
and an EPA Method 5 console equipped with a dry gas meter, a
digital electronic thermometer, and an inclined manometer. Reeve
Angel Type 934 AH substrate was used on each stage of the Ander-
son sampler.
52
-------
SECTION 6
SAMPLING AND ANALYTICAL PROCEDURES
SUMMARY
The general sampling and analytical procedures used and the num-
ber of tests performed are described in Section 2 of this report.
Particulate emissions downstream of the two-stage mechanical
collector were monitored by EPA Methods 5 and 5B. Particulate
emissions upstream were measured by Method 5B. The flue gas
velocity, temperature, flow rate, oxygen content, and carbon
dioxide content were measured by EPA Methods 1-4. The sulfuric
acid and sulfate contents of the particulates collected were
measured by techniques supplied to MRC by EPA on October 10,
1980. The plume opacity was measured by EPA Method 9. Grab
samples of coal were taken before, during, and after every Meth-
od 5 and 5B emission test and analyzed by ASTM D3176 ultimate
analysis, ASTM D2015-66(72) heating value analysis, and a modi-
fied version of ASTM D410-38 sieve size analysis. Further details
on the sampling, analysis, and quality assurance procedures are
given below.
FIELD SAMPLING
While the Method 5 and 5B testing during each run was approxi-
mately simultaneous, a decision was made by EPA not to shut down
all sampling trains when sampling ports were changed. This
resulted in small deviations from true simultaneous sampling at
the exhaust versus the inlet. The times that samples were taken,
filters changed, etc., are provided in the field data sheets in
Appendix A-4.
53
-------
Each Andersen sampler run was conducted for 3 minutes at the
inlet or 20 minutes at the exhaust under isokinetic conditions.
After the completion of each run, the Andersen heads were opened
and the substrates inspected and stored for weighing at MRC's
laboratory.
Grab samples of coal were taken with a shovel through all three
sampling ports at the bottom of the coal feed chute and stored in
1.9 x 10~3 m3 (% gallon) polypropylene containers.
SAMPLE HANDLING
Method 5, Method 5B, and Andersen impactor filters were trans-
ferred to closed containers after sampling. Deposits on the
inside of the sampling equipment were removed with acetone and
bottled. The equipment was then rinsed with distilled water, but
the water rinses were discarded. Access to the samples was
limited in the field; they were stored in locked MRC vehicles
except when they were being handled by authorized individuals.
The samples were shipped to the MRC Dayton Laboratory for analy-
sis in the vehicles used for storage in the field. MRC's usual
security procedures were used to limit access to the samples
while they were being analyzed, and chain-of-custody records were
maintained.
CHEMICAL ANALYSIS
Details of the procedure used to analyze the sulfate and sulfuric
acid content of the Method 5 and 5B stack samples are summarized
in Appendix E. It is based on the procedure supplied to MRC by
EPA at a meeting on October 10, 1980.
After the probe rinses and filters were dried and weighed, room
temperature isopropanol was added to each sample. The samples
soaked for at least 12 hours, then the filter in isopropanol was
54
-------
ultrasonically extracted for 30 minutes. The extracts were
filtered and analyzed for sulfuric acid using the barium-thorin
titration. This isopropanol extraction and analysis procedure
was performed twice on every sample. Those portions of the
extracts that were not titrated were bottled and retained.
After the sulfuric acid extraction with isopropanol, water was
added to the filters and solid residues. After they soaked for
at least 12 hours, they were ultrasonically extracted for 30
minutes. The extracts were filtered, passed through a Rexyn-101
ion exchange column, and analyzed for sulfates (using barium-
thorin titration). The water extraction and sulfate analysis
procedure was performed twice on every sample, and the untitrated
water extracts were bottled and retained.
After all extractions, the filters and solid residues were dried
and saved under dry nitrogen in a refrigerator, along with the
untitrated extracts. These samples will be saved for no more
than 6 months.
Blank filters and residue from the evaporation of clean acetone
were analyzed when the stack samples were analyzed, along with
standards and quality assurance audit samples of sulfate and
sulfuric acid.
No chemical analysis was performed for the Andersen cascade
impactor samples; only particulate masses were determined.
COAL ANALYSIS
Grab samples of the coal being fed into the boiler were taken
before, during, and after every Method 5 and 5B stack sample run.
These coal samples were split in half using ASTM D2013 riffle
sampler protocol, and appropriate half samples were combined to
make one aggregate coal sample for every Method 5/5B sample run.
55
-------
The carbon, hydrogen, nitrogen, oxygen, sulfur, ash, and moisture
content of the coal for each Method 5 and 5B sampling run was
measured using ASTM D3176 ultimate analysis. The energy content
(Btu) was measured by ASTM D2015-66(72) bomb calorimetry, and the
size distribution of each aggregate sample was measured using a
modified version of ASTM D410-38 sieve analysis. The coal samples
were not large enough to meet all applicable ASTM sieve analysis
requirements for all size ranges, and a 2.54 cm (1 in.) screen
was used to separate large pieces of coal from the finer mater-
ial, which was sifted through #4, #16, #30, and #200 sieves.
DATA REDUCTION
MRC's computer and programmable calculators were used to reduce
the analytical and field data to determine results. The "F"
values used to determine ng/ joule (lb/106 Btu) emissions were
taken from the boiler emission regulations in, 40 CFR 60.45(f)
Appendix A contains copies of all raw field data sheets (except
for opacity monitoring) and coding sheets for data processing.
Appendix B contains complete printouts of the results of the
sampling.
Appendix C contains the opacity monitoring results .
Appendix D contains boiler monitoring data taken during the
testing.
Appendix E contains the detailed analytical methods used.
Appendix F contains sample analysis data and a summary of the
results of the quality control and assurance procedures.
56
-------
Appendix G identifies the people performing the sampling, analy-
sis, and data reduction.
QUALITY ASSURANCE
The quality assurance and control program included all applicable
procedures specified in the Federal Register for EPA Methods 1
through 5 and the procedures specified in the EPA Guidelines for
the Development of Quality Assurance Programs for these methods.
The pitot tube on the special stack test system for Method 5 and
5B sampling through the same port was calibrated with the Meth-
od 5 and 5B probes and nozzles in place. This was the same
configuration used during the emission tests.
Particle deposits on the stages of the Andersen sampler were
checked to assure that the stages remained tight during saimpling.
A certified smoke inspector was used for opacity monitoring by
EPA Method 9.
MRC's Quality Assurance and Quality Control (QA/QC) Supervisor
obtained sulfate quality assurance audit filters from the Quality
Assurance Division of the.U.S. Environmental Protection Agency.
Sulfuric acid audit samples were prepared at the Dayton Labora-
tory by the MRC QA/QC Supervisor. The contents of the audit
samples were known only to the QA/QC Supervisor, not to the chem-
ists and technicians performing the analyses. Four blank filters
and sulfate and sulfuric acid quality assurance audit samples
were analyzed along with the field samples. Results of QA/QC
analyses are presented in Appendix F-4.
Standard ASTM procedures were used for the coal analysis. To
obtain an indication of the precision of the coal analysis data,
half of a raw coal sample that was previously analyzed as part of
a November 1980 emission test at the General Motors Chevrolet
57
-------
plant in Parma, Ohio (relabeled as a Parkersburg sample) was
analyzed with the Parkersburg samples. Because it was a leftover
half sample, the Parma sample was not split in half prior to
analysis.
This was the only change from the procedure used for the actual
Parkersburg samples. The chemists analyzing the samples were not
told that this sample was part of a sample analyzed earlier.
After the results of the analysis of the other half of the raw
coal sample previously analyzed were reported, the analysts were
informed of the fact that this sample was analyzed previously.
Portions of the coal sample halves that were homogenized, pul-
verized, and prepared for analysis were then analyzed a second
time for sulfur content and ash content. The purpose of this was
to help determine if the small sample size and sample preparation
procedures may have been a significant cause of the poor reprodu-
cibility of the sulfur and ash content results observed, or if
this was caused mainly by error in the chemical analysis. The
results (reported in Table 9 of Section 2) suggested that the
small sizes of the samples and real differences in the composi-
tion of coal from one piece to another were the main causes of
the poor reproducibility of the results of the analysis of dif-
ferent "halves" of the same raw coal samples.
The data used in computerized data processing were checked by
comparing the printout of the data used to calculate results with
the raw field data used to code the computer input.
58
------- |