oEPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 80-IBR-11
November 1980
Air
Industrial Boilers
Emission Test Report
General Motors
Corporation
Parma, Ohio
-------
INDUSTRIAL BOILERS
Emission Test Report
General Motors Corporation
Parma, Ohio
November 10-17, 1980
Technical Directive No. 4
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
(ESED 80/33)
November 1980
MONSANTO RESEARCH CORPORATION
DAYTON LABORATORY
Dayton, Ohio 45407
-------
CONTENTS
Page
Figures iv
Tables v
1. Introduction
2. Summary of Results 3
Description of Monitoring 3
Summary of Results 5
3. Process Description and Monitoring 25
Process Description 25
Monitoring of Plant Processes During Testing ... 29
Comments on Tests 38
4. Location of Sampling Points 41
Scrubber Inlet 41
Scrubber Exhaust 43
Coal Sampling Site 43
5. Description of Sampling Trains 45
Particulate Sampling Trains 45
Particulate Size Distribution Sampling Apparatus . 45
6. Sampling and Analytical Procedures 49
Summary 49
Field Sampling 49
Sample Handling 50
Chemical Analysis 50
Coal Analysis 51
Data Reduction 52
Quality Assurance 53
LCC/D iii
-------
CONTENTS (continued)
Page
Appendices (continued)
E. Analytical Methods for Sulfuric Acid and
Sulfate Determinations E.I
F. Analytical Data F.I
G. Project Participants G.I
IV
-------
FIGURES
Number Page
1 Cumulative size distribution for first inlet
Andersen run at General Motors Chevrolet Plant
in Parma, Ohio 18
2 Cumulative size distribution for second inlet
Andersen run at General Motors Chevrolet Plant
in Parma, Ohio 19
3. Cumulative size distribution for scrubber exhaust
Andersen run at General Motors Chevrolet Plant
in Parma, Ohio 20
4 Schematic diagram of Boiler No. 1 and associated
equipment at General Motors Chevrolet Plant in
Parma, Ohio 26
5 Flow diagram of double-alkali scrubbing unit. . 30
6 Scrubber inlet sampling site at Parma, Ohio
GM Plant 42
7 Scrubber exhaust sampling site at Parma, Ohio
GM Plant 44
8 Particulate sampling train - EPA Methods 5
and 5B 46
9 Particulate size distribution sampling
apparatus - Andersen 2000, Inc 47
-------
TABLES
Number Page
1 Source Sampling and Analyses at Chevrolet Plant
in Parma, Ohio (November 10-17, 1980) 4
2 Emission Data and Stack Gas Parameters, General
Motors Chevrolet Plant, Parma, Ohio,
November 10-17, 1980 (Metric Units) 6
3 Emission Data and Stack Gas Parameters, General
Motors Chevrolet Plant, Parma, Ohio,
November 10-17, 1980 (English Units) 9
4 Summary of Scrubber Efficiency Measurements
and Differences between Method 5 and Method 5B
Results at General Motors Chevrolet Plant in
Parma, Ohio (November 10-17, 1980) 13
5 Summary of Duration of Sampling, Stack Temperature,
Stack Flow Rate, Sample Volume, and Water Content
of Methods 5 and 5B Stack Samples at the General
Motors Chevrolet Plant, Parma, Ohio
(November 10-17, 1980) 15
6 Summary of Integrated Gas Analysis Results at
the General Motors Chevrolet Plant, Parma,
Ohio (November 10-17, 1980) 16
7 Summary of Andersen Particle Sizing Results at
the General Motors Chevrolet Plant, Parma,
Ohio (November 10-17, 1980) 17
8 Summary of Plume Opacity Observations at the
General Motors Chevrolet Plant in Parma, Ohio
(November 10-17, 1980) 21
9 Summary of Coal Analysis at the General Motors
Chevrolet Plant, Parma, Ohio (November 10-17,
1980) 23
10 Design, Operating, and Performance Characteristics
of Double-Alkali Scrubbing System 27
VI
-------
TABLES (continued)
Number Page
11 Power Plant Operating Parameters (11/12/80,
Run SI) 32
12 Power Plant Operating Parameters (11/13/80,
Run 1) 33
13 Power Plant Operating Parameters (11/13/80,
Run 2) 34
14 Power Plant Operating Parameters (11/14/80,
Run 3R) 35
VI1
-------
SECTION 1
INTRODUCTION
Emissions from Boiler No. 1 at the General Motors Chevrolet Plant,
on Brookpark Road in Parma, Ohio, were tested November 11-14, 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 Assign-
ment No. 2. The boiler tested was a 12.6 kg/s (100,000 Ifo/hr)
steam capacity waste oil and coal-fired spreader stoker with an
economizer, a multiclone, and a double-alkali scrubber.
The objective of the sampling program is to provide background
information on well-controlled sources for the development of new
source performance standards. Within this framework, the objec-
tives of the sampling at the Parma, Ohio, General Motors Plant
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 scrubber
(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 double-alkali scrubber on par-
ticulate, sulfate, and sulfuric acid emissions, as measured
by EPA Method 5B.
-------
The field test was monitored by Dennis Holzschuh, Emission Measure-
ment Branch, EPA. The sampling was directed by Richard G. Beer,
MRC team leader.
The monitoring performed at this plant consisted of three simul-
taneous runs of Methods 5 and 5B at the scrubber outlet performed
simultaneously with Method 5B sampling at the scrubber inlet.
Separate stack test trains, traversing on perpendicular diameters
of the stack, were used for the tests at the scrubber outlet. One
additional single-point run of Methods 5 and 5B was performed at
the scrubber outlet with both sampling systems operating simulta-
neously at essentially the same point in the stack and with no
testing at the scrubber inlet. The purpose of the run of Methods
5 and 5B at a single point in the stack was to obtain an indica-
tion of whether or not the differences between the results of
simultaneous Method 5 and 5B sampling are significantly different
when both methods were used at a single point versus when they
were used with two stack test trains traversing the same stack
independently.
The boiler was running under steady state conditions (no soot
blowing, ash unloading, etc.) at a minimum of 87% capacity during
these tests, and the coal burned was considered to be representa-
tive of normal feed.
The filters and dried acetone and water rinses of the Method 5 and
5B runs were weighed and analyzed for sulfuric acid and sulfates
using procedures described in the Sampling and Analytical Proce-
dures section of this report.
The plume opacity was measured during all Method 5 and 5B tests
by EPA Method 9, when possible. In addition, particle size dis-
tributions were measured twice at the inlet to the scrubber and
once at the outlet, using Andersen cascade impactors.
-------
SECTION 2
SUMMARY OF RESULTS
DESCRIPTION OF MONITORING
Table 1 summarizes the monitoring that was performed at this
plant. It consisted of three simultaneous runs of Methods 5 and
5B at the scrubber outlet performed simultaneously with Method 5B
sampling at the scrubber inlet. Separate stack test trains,
traversing on perpendicular diameters of the stack, were used for
the tests at the scrubber outlet. Sample volumes of at least
2.4 dry m3 (85 dry ft3) were taken during these runs. One addi-
tional single-point run of Methods 5 and 5B was performed at the
scrubber outlet, with both sampling systems operating simultaneous-
ly at essentially the same point in the stack, with no testing at
the scrubber inlet. About 1.9 dry m3 (68 ft3) was sampled by each
sampling train during the single point sampling run. The boiler
was running under steady state conditions (no soot blowing, ash
unloading, etc.) at a minimum of 87% of its full capacity during
these tests. The coal burned was considered to be representative
of normal feed.
One additional run of Method 5B at the inlet and Methods 5 and 5B
at the outlet occurred, but it was terminated before the traverses
were completed when it became evident that the coal being burned
contained unusually high amounts of coal dust.
Distilled water was used to remove deposits from the inside of
the Method 5 and 5B probes, nozzles, and filter holders, after
the usual acetone rinses. This was done because, during previous
-------
TABLE 1. SOURCE SAMPLING AND ANALYSES AT CHEVROLET
PLANT IN PARMA, OHIO (NOVEMBER 10-17, 1980)
SAMPLING AND ANALYSIS REQUIREMENTS
MRC Job No. 101.1221
Total
no. of
samples
3*
34
3a
1
1
2
1
4
26b
Sample
type
articulate & ORSAT
crubber InTet
Scrubbet Bxhaust
articulate t ORSAT
crubber Exhaust
articulate & ORSAT
crubber Exhaust
articulate £, ORSAT
crubber Exhaust
ORSAT Scrubber
Inlet
ORSAT Scrubber
Exhaust
Plune Opacity
Coal
Sampling
method
SB
5B
5
iingle Poin
5-
iingle Poin
tndersen
Lnderaen
9C
Crab
Contract No. : bs-02-3547
Company Name: GENERAL MOTORS
Industry: INDUSTRIAL BOILER
Sample
collected
by
Minimum
sampling
time
120 min
20 min
s multaneoi
* th above
0 min
i multaneoi
i th above
60 min
60 min
iimultaneoi
rith above
)epends on
.oading
tepends on
wading
iimultaneoi
ilth Methoi
S i. SB
Before,
during anc
Method S
tests
Assignment Number: WA ,*2 Technical Directive: 2
CHEVROLET Company Location: PARMA, OHIO
Process: COAL t WASTE OIL-FIRED
SPREADER STOKER
Minimum
volume gas
sampled ft3
60
s 60
s 60
60
s 60
Depends on
Loading
Depends on
Loading
s
Not
Applicable
Not
Applicable
Initial Analysis
Type Method By
Velocity
temp, flow
Oa li. C0a
Velocity
temp, flow
Oa t, COa
Velocity
temp, flow
Oa t COa
Velocity
temp, flow
Oj S. COa
Velocity
temp, flow
Oa ' CO2
Velocity
temp.
Oa ' COa
Velocity
temp.
Oa ' COa
Visual
Observatio
Split and
Prepare 8"
aggregate
Samples fo
runs Rl,
R2, R3
R3 retest
L AI-R2,
(PO RS t.
R6)
1-4
1-4
1-4
1-4
1-4
2-4
2-4
9C
iSTH D201
172)
MRC
Control Equipment: MULTICLONES
t DUAL-ALKALI SCRUBBER
Final Analysis
Type Method By
Particulate
50»-J,
HaSO.
Particulate
SO,-",
HaSO«
'articulate
HaSO,
Particulate
H,SO.
HaSO,
'article
iize
pistributio
'article
iice
>istributio
tot
applicable
Iltl
BTU,
mate,
Size
Method S
> 10/10/80
Procedure
Method S
t 10/10/80
Procedure
Method S
t 10/10/80
'rocedure
Method 5
> 10/10/80
Procedure
Method S
f i 0/1 0/80
Procedure
kndersen
n
uidersen
tot
applicable
fkSTM D3176
ASTM D2015
60, 1721
ISTM D410-
38
Except use
ivailable
lample 14,
16, 30, 20
Sieves 6
1" screen.
MRC
One additional run occurred but it was terminated prior to completing the traverse.
19 of these samples were split and combined into aggregate samples for all four
Method 5/5B traverse tests, one of the single point tests and one of the Andersen
tests.
•»
'Not all of these measurements were strictly Method 9 because of an interfering steam
plume.
-------
testing on boilers, yellow films had been observed on the probe
linings of Method 5 trains after acetone rinses. Sulfate deposits
were a suspected cause of the film.
The filters and dried acetone and water washes of the Method 5
and 5B runs were weighed and analyzed for sulfuric acid and
sulfates.
Methods 1 through 4 were used during all Method 5 sampling runs,
as in typical compliance monitoring.
The plume opacity was measured during all Method 5 and 5B tests
by EPA Method 9 when possible. However, a steam plume from
another stack prevented valid Method 9 readings from being made
during much of the emission testing. In addition, particle size
distributions were measured twice at the inlet to the scrubber
and once at the outlet, using Andersen cascade impactors.
SUMMARY OF RESULTS
The particulate, sulfuric acid, and sulfate emissions measured by
Methods 5 and 5B are summarized in Tables 2 and 3. In addition,
Table 3 shows the percents of the concentrations measured by
Method 5 at the outlet represented by the concentrations measured
by Method 5B simultaneously. Run 3 is the partial traverse run,
and run SI is the run using Methods 5 and 5B at a single point in
the scrubber exhaust. The emission rates in Tables 2 and 3 do
not include the water washes of the Method 5 and 5B trains. Par-
ticulate emissions represent all emissions measured by weighing
the Method 5 or 5B samples (including sulfuric acid and sulfates).
Particulate concentrations measured by Method 5B traverses were
50% to 84% of those measured by Method 5. On the single point
run, Method 5B particulate concentrations were 56% of particulates
measured by Method 5. Sulfuric acid concentrations measured on
-------
TABLE 2. EMISSION DATA AND STACK GAS PARAMETERS, GENERAL MOTORS
CHEVROLET PLANT, PARMA, OHIO, NOVEMBER 10-17, 1980
(METRIC UNITS)
Emissions
Run Sampling
number Date Location method Pollutant
1 11-13-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
2 11-13-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
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
g/dscm
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.5165
.0968
.0489
.0164
.0333
.0029
.0264
.0132
.0129
.7981
.1154
.0976
.0159
.0362
.0176
.0236
.0122
.0180
kg/hr
25.8
4.7
2.4
0.8
1.6
0.1
1.3
0.6
0.6
41.1
5.6
4.9
0.8
1.8
0.9
1.2
0.6
0.9
ng/J
256
48
23
8
16
I
13
6
6
385
57
48
1
18
8
11
6
8
.4
.9
.8
.1
.8
.4
.1
.7
.3
.8
.8
.5
.7
.1
.7
.4
.1
.9
Cor-
rec-
ted
to 12%
C02,
g/dscm
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.6324
.1223
.0593
.0201
.0421
.0035
.0323
.0167
.0156
.958
.1473
.1195
.0191
.0462
.0216
.0283
.0156
.0220
Iso-
ki-
netic
% Notes
100
103
99
100
103
99
100
103
99
98
101
98
98
101
98
98
101
98
.5
.0
.2
.5
.0
.2
.5
.0
.2
.9
.3
.4
.9
.3
.4
.9
.3
.4
(continued)
-------
TABLE 2 (continued)
Emissions
Run
number Date Location
3 11-14-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
3R 11-14-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Sampling
method Pollutant
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
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
g/dscm
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.3336
.1990
.2897
.0062
.0260
.0095
.0197
.0198
.0274
.5646
.1016
.0598
.0020
.0325
.0068
.0293
.0153
.0141
kg/hr
12
9
13
0
1
0
0
0
1
6
4
2
0
1
0
1
0
0
.7
.3
.4
.2
.2
.4
.8
.9
.3
.8
.8
.8
.1
.5
.3
.4
.7
.7
ng/J
162
98
147
3
12
4
9
9
14
261
49
31
0
15
3
13
7
7
.7
.8
.8
.0
.9
.8
.6
.8
.0
.5
.6
.1
.9
.9
.5
.6
.5
.3
Cor-
rec-
ted
to 12%
C02,
g/dscm
0.4170
0.2437
0.3698
0.0078
0.0318
0.0121
0.0246
0.0242
0.0350
0.6392
0.1219
0.0772
0.0023
0.0390
0.0088
0.0332
0.0184
0.0182
Iso-
ki-
netic
%
105.4
99.5
102.0
105.4
99.5
102.0
105.4
99.5
102.0
101.3
99.4
100.1
101.3
99.4
100.1
101.3
99.4
100.1
Notes
Run terminated
prior to com-
pleting trav-
erses because
excessive
amounts of
coal dust
being burned
(continued)
-------
TABLE 2 (continued)
Emissions
Run
number Date Location
SI 11-12-80 Outlet
Outlet
Outlet
Outlet
Outlet
Outlet
Sampling
method Pollutant
5
5B
5
5B
5
5B
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Acutal
g/dscm
0
0
0
0
0
0
.0955
.0536
.0292
.0026
.0152
.0033
kg/hr
5
2
1
0
0
0
.0
.6
.5
.1
.8
.2
ng/J
48
25
14
1
7
1
.3
.7
.8
.2
.7
.6
Cor-
rec-
ted
to 12%
C02/
g/dscm
0.1169
0.0624
0.0358
0.0033
0.0186
0.0038
Iso-
ki-
netic
% Notes
96
91
96
91
96
91
.5
.9
.5
.9
.5
.9
00
-------
TABLE 3. EMISSION DATA AND STACK GAS PARAMETERS, GENERAL MOTORS
CHEVROLET PLANT, PARMA, OHIO NOVEMBER 10-17, 1980
(ENGLISH UNITS)
Emissions
Run
number Date Location
1 11-13-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
2 11-13-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Sampling
method Pollutant
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
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
gr/dscf
0.2256
0.0423
0.0213
0.0072
0.0145
0.0013
0.0115
0.0058
0.0057
0.3487
0.0504
0.0426
0.0069
0.0158
0.0077
0.0103
0.0053
0.0079
Ib/hr
56.8
10.4
5.3
1.8
3.6
0.3
2.9
1.4
1.4
90.6
12.4
10.8
1.8
3.9
1.9
2.7
1.3
2.0
Ib/mm
Btu
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.596
.114
.055
.019
.039
.003
.030
.016
.015
.897
.134
.113
.018
.042
.020
.026
.014
.021
Correc-
ted to
12% C02,
gr/dscf
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.2762
.0534
.0258
.0088
.0183
.0016
.0141
.0073
.0069
.4184
.0643
.0522
.0083
.0202
.0094
.0124
.0068
.0097
Percent of
Method 5
Iso- g/dscf repre-
ki- sented by
netic Method 5B g/
J% dscf at outlet
100.
103.
99.
100.
103.
99.
100.
103.
99.
98.
101.
98.
98.
101.
98.
98.
101.
98.
5
0
2
5
0
2
5
0
2
9
3
4
9
3
4
9
3
4
50.4
9.0
98.3
84.5
48.7
149.1
(continued)
-------
TABLE 3 (continued)
Emissions
Run
number Date Location
3 11-14-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
3R 11-14-80 Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Sampling Actual
method Pollutant gr/dscf
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
Particulate 0.14573
Particulate 0.08703
Particulate 0.1266a
Sulfuric acid 0.0027
Sulfuric acid 0.0114
Sulfuric acid 0.0041
Sulfates 0.0086
Sulfates 0.0087
Sulfates 0.0120
Particulate 0.2467
Particulate 0.0444
Particulate 0.0261
Sulfuric acid 0.0009
Sulfuric acid 0.0142
Sulfuric acid 0.0030
Sulfates 0.0128
Sulfates 0.0067
Sulfates 0.0061
Ib/mm
Ib/hr Btu
0.3783
0.230a
0.3443
0.007
0.030
0.071
0.022
0.023
0.033
59.1 0.608
10.6 0.115
6.3 0.072
0.2 0.002
3.4 0.037
0.7 0.008
3.1 0.032
1.6 0.017
1.5 0.017
Correc-
ted to
12% C02,
gr/dscf
0.18213
0.10653
0.16163
0.0034
0.0140
0.0052
0.0108
0.0107
0.0153
0.2793
0.0533
0.0337
0.0010
0.0170
0.0039
0.0145
0.0080
0.0079
Percent of
Method 5
Iso- g/dscf repre-
ki- sented by
netic Method 5B g/
% dscf at outlet
105.4
99.5
102.0
105.4
99.5
102.0
105.4
99.5
102.0
101.3
99.4
100.1
101.3
99.4
100.1
101.3
99.4
100.1
a
36.0
137.9
58.8
21.1
91.0
(continued)
-------
TABLE 3 (continued)
Emissions
Run
number Date Location
SI 11-12-80 Outlet
Outlet
Outlet
Outlet
Outlet
Outlet
Sampling
method Pollutant
5
5B
5
5B
5
5B
Particulate
Particulate
Sulfuric acid
Sulfuric acid
Sulfates
Sulfates
Actual
gr/dscf
0.0417
0.0234
0.0128
0.0011
0.0066
0.0014
Ib/hr
11.0
5.8
3.4
0.3
1.7
0.4
Ib/mm
Btu
0
0
0
0
0
0
.112
.060
.034
.003
.018
.004
Correc-
ted to
12% C02,
gr/dscf
0
0
0
0
0
0
.0511
.0273
.0157
.0013
.0081
.0016
Percent of
Method 5
Iso- g/dscf repre-
ki- sented by
netic Method 5B g/
% dscf at outlet
96
91
96
91
96
91
.5
.9
.5
.9
.5
.9
56.1
8.6
21.2
Run terminated prior to completing traverses because excessive amounts of coal dust being burned.
-------
the Method 5B traverses were 9% to 49% of the sulfuric acid con-
centrations on the Method 5 samples (it was 9% for the single
point run). Sulfate concentrations measured by Method 5B tra-
verses were 91% to 121% of the concentrations by Method 5 (it
was 21% for the single point run).
The weights of particulate collected in the water washes were
less than 3% of the weights of the particulate in the filters and
acetone rinses. Appendix A. 3 shows the weights of particulate,
sulfuric acid, and sulfates in the filters, acetone rinses and
water rinses. Appendix A.3 also cross references the run numbers
in this section with the run numbers in the appendices and on the
computer printouts. Computer printouts of the particulate,
sulfuric acid, and sulfate emission rates associated with the
filters, acetone rinses and water rinses are provided in
Appendix B.
Table 4 summarizes the particulate, sulfate, and sulfuric acid
collection efficiency of the scrubber, as measured by EPA Method
5B. The differences between emissions measured by Method 5 and
5B are also summarized in this table. Finally, this table
summarizes what percent of the differences between Method 5 and
5B particulate emissions can be accounted for by differences
in the amounts of sulfuric acid (and sulfuric acid and sulfates
combined) measured by Method 5 versus 5B. Run 3 is not included
(because it was terminated prior to completing the traverse).
Forty-nine to 125% of the amount of difference in particulate
emissions measured by Method 5 versus 5B traverses are accounted
for by differences in the amounts of measured sulfuric acid emis-
sions. For the single point run,, sulfuric acid accounted for 60%
of the particulates. Sulfuric acid and sulfates combined account
for 65% to 82% of the difference in particulate emissions between
measurements by Method 5 and Method 5B traverses. For the single
12
-------
TABLE 4. SUMMARY OF SCRUBBER EFFICIENCY MEASUREMENTS AND DIFFERENCES BETWEEN METHOD
AND METHOD 5B RESULTS AT GENERAL MOTORS CHEVROLET PLANT IN PARMA, OHIO
(NOVEMBER 10-17, 1980)
Run
number
1
2
3R
SI
Percent of difference
Scrubber collection Method 5 in particulate emis-
efficiency, % minus sions kg/hr (Ib/hr)
based Method SB represented by dif-
on kg/hr (Ib/hr) emissions ference in sulfuric
Pollutant emissions kg/hr Ib/hr acid emissions
Particulate
Sulfuric acid
Sulfates
Sulfuric acid
and sulfates
Particulate
Sulfuric acid
Sulfates
Sulfuric acid
and sulfates
Particulate
Sulfuric acid
Sulfates
Sulfuric acid
and sulfates
Particulate
Sulfuric acid
Sulfates
91%
83%
52%
88%
6%
26%
89%
-250%
52%
2.3
1.5
0.0
1.5
0.7
0.9
-0.3
0.6
2.0
1.2
0.0
1.2
2.4
1.4
0.6
5.1
3.3 64.7
0.0
3.3
1.6
2.0 125.0
-0.7
1.3
4.3
2.1 48.8
0.1
2.8
5.2
3.1 59.6
1.3
Percent of difference
in particulate emis-
sions kg/hr (Ib/hr)
represented by dif-
ference in sulfuric
acid and sulfates
combined
64.7
81.3
65.1
Sulfuric acid
and sulfates
2.0
4.4
84.6
-------
point run, sulfuric acid and sulfates combined accounted for 85%
of the particulates.
Table 5 summarizes the duration of the Method 5 and 5B sampling,
the sample volumes, the stack temperatures, the stack flow rates,
and the water content of the flue gas. Table 6 summarizes the
flue gas carbon dioxide, carbon monoxide, oxygen and nitrogen
measurements associated with the Method 5 and 5B samples. Differ-
ences between the results associated with the Method 5 and 5B
sampling at the outlet in Tables 5 and 6 are indicative of the
precision of the measurements.
Table 7 summarizes the results of the Andersen cascade impactor
sampling. The cumulative aerodynamic size distributions are
graphed in Figures 1, 2, and 3.
Table 8 summarizes the results of the plume opacity observations
by EPA Method 9. This data should be viewed with caution since the
opacity observations were taken about 15 m (50 ft) downwind of the
mouth of the stack where the scrubber steam plume dissipated; and
visible plumes from other stacks sometimes mixed with the plume from
Boiler No. 1 before the steam dissipated, preventing the opacity
from being measured during portions of the observation periods.
Table 9 summarizes the results of the coal sampling and analysis.
To obtain an indication of the precision of the coal analysis
data, half of the raw coal sample taken during the second Andersen
run at the inlet at Parma was analyzed along with a set of coal
samples from an emission test in Parkersburg, West Virginia. 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 (2.62%
versus 1.35% dry) and in the ash content (8.05% versus 3.92% dry).
Portions of the same half-samples discussed above were then anal-
yzed a second time; the differences in sulfur content were between
14
-------
TABLE 5. SUMMARY OF DURATION OF SAMPLING, STACK TEMPERATURE, STACK FLOW RATE, SAMPLE
VOLUME, AND WATER CONTENT OF METHODS 5 AND 5B STACK SAMPLES AT THE GENERAL
MOTORS CHEVROLET PLANT, PARMA, OHIO (NOVEMBER 10-17, 1980)
Run
1
2
3
3R
SI
Location
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Outlet
Outlet
Sampling
method
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5
5B
Duration
of
sampling,
min
141
144
144
144
144
144
60
60
60
144
144
144
108
108
Measured
stack
temperature
°C °F
167
47
48
176
47
48
159
47
48
165
47
47
48
48
332
117
118
349
117
118
318
117
118
328
116
117
118
118
Stack flow rate
dscm/
min
832
813
827
859
815
834
636
778
774
792
789
792
872
815
dscf/
min
29,361
28,696
29,185
30,332
28,780
29,438
22,443
27,469
27,326
27,966
27,843
27,953
30,802
28,792
Sample
dscm
2.52
2.57
2.50
2.62
2.54
2.50
0.86
0.99
1.00
2.47
2.41
2.41
1.94
1.71
volume
dscf
89.03
90.86
88.23
92.50
89.56
88.32
30.37
34.99
35.38
87.34
85.07
85.25
68.55
68.46
Water
content,
%
6.04
11.68
11.69
5.93
14.31
13.08
11.13
12.12
12.67
6.84
12.04
12.56
11.69
12.18
-------
TABLE 6. SUMMARY OF INTEGRATED GAS ANALYSIS RESULTS AT THE
GENERAL MOTORS CHEVROLET PLANT, PARMA, OHIO
(NOVEMBER 10-17, 1980)
Run
1
2
3
3R
SI
Location
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Inlet
Outlet
Outlet
Outlet
Outlet
Sampling
method
before
Method 3
train
5B
5
5B
5B
5
5B
5B
5
5B
5B
5
5B
5
5B
C02/ %
9.8
9.5
9.9
10.0
9.4
9.8
9.6
9.8
9.4
10.6
10.0
9.3
9.8
10.3
CO, % 02, %
<0.1 9.8
<0.1 10.0
<0.1 9.6
<0.1 9.5
<0.1 9.9
<0.1 9.8
<0.1 9.6
<0.1 9.8
<0.1 10.1
<0.1 9.0
<0.1 9.6
<0.1 10.3
<0.1 10.0
<0.1 9.4
N2, %
80.4
80.5
80.5
80.5
80.7
80.4
80.8
80.4
80.5
80.4
80.4
80.4
80.2
80.3
Dry
molecular
weight,
kg/kg mole
(Ib/lb mole)
30.0
29.9
30.0
30.0
29.9
30.0
29.9
30.0
29.9
30.1
30.0
29.9
30.0
30.0
16
-------
TABLE 7. SUMMARY OF ANDERSEN PARTICLE SIZING RESULTS AT THE GENERAL
MOTORS CHEVROLET PLANT, PARMA, OHIO (NOVEMBER 10-17, 1980)
Flow rate Percent
Location Run acm/min acf/min isokinetic Stage
Inlet 1 0.011 0.40 103.4 0
1
2
3
4
5
6
7
Final filter
Inlet 2 0.012 0.41 106.0 0
1
2
3
4
5
6
7
Final filter
Outlet 1 0.017 0.60 101.2 0
1
2
3
4
5
6
•~i
i
Final filter
Size range,
pm
>15
10.6 -
7.3 -
4.9 -
3.15 -
1.56 -
0.917 -
0.67 -
0 -
>15
10.5 -
7.1 -
4.8 -
3.05 -
1.55 -
0.95 -
0.66 -
0 -
>11
7.7 -
5.2 -
3.6 -
2.25 -
1.12 -
0.70 -
0.47 -
0 -
.6
15.6
10.6
7.3
4.9
3.15
1.56
0.97
0.67
.4
15.4
10.5
7.1
4.8
3.05
1.55
0.95
0.66
.5
11.5
7.7
5.2
3.6
2.25
1.12
0.70
0.47
Percent
in size
range
78.8
3.52
0.70
0
1.41
10.56
1.41
0.70
2.82
76.92
1.22
6.88
0
0.81
4.86
0
3.64
5.67
8.76
0
0
4.38
4.38
24.08
16.06
21.90
20.44
Cumulative
percent,
-------
00
0 t0 30 y.o r.o »• o l.o
PARTICLE AERODYNAMIC DIAMETER - pm
16.
Figure 1. Cumulative size distribution for first inlet Andersen
run at General Motors Chevrolet Plant in Parma, Ohio
(November 12, 1980).
-------
L.» 3 0 y.o f.o ».« 10 '••*'„
PARTICLE AERODYNAMIC DIAMETER
Figure 2. Cumulative size distribution for second inlet Andersen
run at General Motors Chevrolet Plant in Parma, Ohio
(November 13, 1980).
-------
PARTICLE AERODYNAMIC DIAMETER - \im
Figure 3. Cumulative size distribution for scrubber exhaust Andersen
run at General Motors Chevrolet Plant in Parma, Ohio
(November 14, 1980).
-------
TABLE 8. SUMMARY OF PLUME OPACITY OBSERVATIONS AT THE
GENERAL MOTORS CHEVROLET PLANT IN PARMA, OHIO
(NOVEMBER 10-17, 1980)
Date: 11/11/80
Type of Discharge: Stack
Height of Point of Discharge: 100 ft
Wind Direction: N
Color of Plume: White
Observer Name: C Clark
Distance from Observer
to Discharge Point: 250 ft
Direction of Observer
from Discharge Point: SW
Height of Observation Point: Ground Level
Description of Background: Water tower
Distance Downwind of Stack where
Steam Plume Dissipated and
Plume Opacity Measured: 50 ft
Type of Plant: Industrial Boiler
Location of
Discharge: Scrubber exhaust
Description of Sky: Overcast
Wind Velocity: 7-10 mph
Duration of Observation: 100 min
o
13:
14:
14:
:30
:36
:42
:48
•54
;00
:06
:12
•IB
?4
•30
;3ft
:42
•4fi
:54
:00
:06
End
12:
12:
12:
12:
1?;
13:
13:
13;
13'
13;
13'
13;
13:
13'
13:
14:
14:
:35
:41
;47
:53
;59
:OS
:11
:17
?3
?9
;35
:41
:47
•S3
:59
:05
:09
Sum
260
275
220
165
240
200
195
120
145
190
220
95
285
240
200
?6B
180
Opacity
Average
all sets
10.
11,
11 .
10,
10
10,
10,
10,
10
10
10
10
1,3
1?
12,
11
12.
11
.8
•s..
.od
.3d
.0
V
,ftd
,od
7d
6d
Od
.6"
,0d
nd
.5"
,n
.9d
.1%
Interference from a steam plume from
another stack prevented the opacity
from being measured during portions of
this six minute time interval.
1 2
TIME, hours
Date: 11/12/80
A plume from Stack #4 was mixed with the plume from the source being
tested and prevented obtaining valid opacity readings.
21
-------
TABLE 8 (continued)
Date: 11/34/80
Type of Discharge: Stack
Height of Point of Discharge: 100 ft
Wind Direction: SW
Color of Plume: White, black for
last 41 min
Distance from Observer
to Discharge Point: 1,050 ft
Direction of Observer
from Discharge Point: SE
Height of Observation Point: 100 ft
Description of Background: Water tower
Distance Downwind of Stack where
Steam Flume Dissipated and
Plume Opacity Measured: 50 ft
Type of Plant: Industrial Boiler
Location of
Discharge: Scrubber exhaust
Description of Sky: Partly cloudy
Wind Velocity: 10-12 mph
Duration of Observation: 251 min
Summary of average opacity
Set
Time
number Start
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
9;
9:
9:
9:
9:
9:
Q ,
9:
9:
9:
10:
10:
10:
10;
10:
10:
10:
10:
10:
10:
11;
:00
:06
;12
24
:30
:36
:42
48
:54
:00
;06
:18
^24
:30
:36
:42
:48
:54
:00
End
9:
9:
9;
9:
9:
9:
9:
9;
9:
9:
10;
10;
10:
10
10:
10:
10:
10
10
10:
11
;05
: 11
:17
123
:29
:35
^47
:53
:59
:05
: 17
^23
:29
:35
^47
:53
:59
:05
Opacity
Sum
170a
120a
185a
80a
200f
210a
200a
225a
135a
225a
215a
195a
115a
130a
205a
120a
195a
180a
230a
160a
170a
Average
9.
10,
9.
10.
9.
10,
10.
9.
10.
11.
9.
9.
7.
9.
9.
7.
9,
9,
10,
10,
7,
4a
oa
',7a
,oa
,5a
•°a
,0a
,8a
,4a
.3a
,8a
,8a
.7a
,3a
,8a
1
'.8*
.5a
.oa
,0a
,7a
Summary of average opacity
Set
number
22
23
24
25
26
27
28
29
30
31
32
33
34
36b
37
38
39
40
41
42C
Time
Start
11;
11:
11:
11:
11:
11:
11;
11:
11:
12
12:
12;
12:
12:
12;
12:
12
12
12
13:
13
:06
. 12
:18
:24
;30
:36
:42
:48
:54
:00
:06
: 18
•24
;3ob
:36
:42
:48
:54
:00
:06
End
111
Hi
11:
11:
11;
11:
11:
11:
11:
12:
12:
12:
12:
12:
12:
12:
12:
12:
12;
13;
13:
ill
: 17
:23'
:29
:35
;41
i47
: 53
:59
:05
: 11
:17
:23
:35
:41
:47
:53
:59
:05
Opacity
Sum
60a
130a
150a
160a
120a
250a
220a
165a
60a
90a
145a
130a
135a
165a
545a
325a
420a
395a
380a
710a
545C
Average
5,
5,
6.
7,
5,
12
12,
9,
10
11
12.
10,
11
11
23
16
23
24
17
30
27
a
. 9a
,5a
•7a
.5a
.2a
•7a
.3a
la
:sa
,3a
.8a
*7
'.3a
.3a
.7a
.3a
.9a
.3C
Average, all sets
11.£
35
30
20
10
Interference from a steam plume from another stack prevented the opacity
from being measured during portions of this six minute time interval.
Plume color changed to black at this time.
cFinal set from a five minute observation period.
RUN T-3
t
I—I
o
0
1
TIME, hours
Date; 11/14/80
Plumes from two other stacks were mixed with the plume from the source being
tested and prevented obtaining of valid opacity readings.
22
-------
KJ
00
TABLE 9. SUMMARY OF COAL ANALYSIS AT THE GENERAL
PARMA, OHIO (NOVEMBER 10-17, 1980)
MOTORS CHEVROLET PLANT,
Parameter
Total mo i stut'e, %
Heating value J/g (Btu/lb)
Dry basis
As received
Ash, %
Dry basis
As received
Sulfur, %
Dry basis
As received
Carbon. %
Dry basis
Hydrogen, %
Dry basis
As received
Oxygen, %
Dry basis
As received
Nitrogen, %
Dry basis
As received
Sieve sample weight, g (Ib)
Percent by weight collected
25. 4 mm sieve (1.0 in.)
«4 sieve, 4.75 era (0.187 in.)
816 sieve, 1.18 nun (0.046 in.)
#30 sieve, 0.60 mm (0.024 in.)
S200 sieve, 0.075 mm (0.003 in.)
Passing through all sieves
Cumulative percent by weight
less than sieve sizes
25 .4 nun (I in. sieve )
«4 sieve. 4.75 mm (0.187 in.)
HIA sieve, 1.19 ss {0 04* ;" '
»30 sieve, 0.60 mm (0.024 in.)
H200 sieve, 0.075 mm (0.003 in.)
Ruti 1
3.
32,
(14,
31,
(13,
5.
4.
1.
1.
79.
75 ,
6.
5.
8.
8.
1.
1.
15
900
100)
800
700)
13
97
09
05
6
.0
6
.6
1
.59
50
2,019.3
(4.
22.
73.
2.
0.
1.
0.
77.
4.
i
1.
0.
45)
46
01
49
41
14
49
54
53
nji
.63
.49
Run 2
3.
33,
(14,
31,
(13,
4.
4
1.
1.
79.
74
5
5
9
8
1
I
66
000
200)
800
700)
.43
.26
.48
.43
,9
. 9
.6
.2
.1
.6
.37
.29
1,981. B
(4
11
79.
5
l!
2.
0
88
8.
3
2
0
.37)
.40
.79
.14
.10
.55
.0
.60
.79
. 55
.55
.0
Run 3
4.
31.
(13,
30,
(13.
7.
7.
1.
1.
76.
70 .
5.
5.
8.
8.
1.
1.
63
BOO
700)
400
100)
40
06
91
82
7
9
5
1
7
1
36
26
1,200.7
(2.
4.
75.
11.
3.
4.
0.
95.
19.
T
4.
0.
65)
61
55
95
23
48
18
39
84
oa
66
18
Run 3R
3.74
33.000
(14,200)
31,700
(13.700)
3.51
3.38
1.43
1.38
81.3
75.9
6.0
5.6
8.5
7.9
1.58
1.47
2,487.5
(5.48)
14.58
82.46
2.03
0.26
0.67
0.20
85.42
3.16
1 11
-------
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 was 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.
Two quality control samples of sulfuric acid were prepared from
0.02 N sulfuric acid and analyzed along with the Parma stack
samples. The measured amounts of sulfuric acid were 101% and 121%
of the reported "true" values. The results of the analysis of two
EPA sulfate quality control filter strip samples (with more than
2 mg of sulfate) showed the measured values to be 87.5% and 100%
of the "true" amounts of sulfate. For two filter strips with
less than 1.0 mg/sample, the percentages were 181% and 195%. The
possibility exists of differences in the physical and chemical
composition of the stack samples (compared to those of the qual-
ity control samples) resulting in differences in the accuracy of
analysis of the stack samples from that reported for the quality
control samples.
Appendix C contains the Method 9 plume opacity measurements. These
readings were taken downwind of the mouth of the stack at locations
after the scrubber water plume dissipated.
24
-------
SECTION 3
PROCESS DESCRIPTION AND MONITORING
PROCESS DESCRIPTION
Boiler System
The steam plant at CMC contains four boilers rated at a total
combined steam generating capacity of 145,000 kg/hr (320,000 Ib/hr),
two with a nominal capacity of 45,000 kg/hr (100,000 Ib/hr) and
two with a nominal capacity of 27,000 kg/hr (60,000 Ib/hr).
Figure 4 shows a flow diagram of a typical boiler and controls.
Each is fired by a spreader stoker with traveling grates and opera-
tes with variable excess air rates in the 100% range. The larger
boilers (1 and 2) are equipped with economizers that lower flue
gas temperature to 135°C (275°F), and the smaller boilers (3 and 4)
operate at a flue gas temperature of 302°C (575°F). Each boiler
is fitted with mechanical dust collectors (multiclones) for primary
particulate control. Normal burning of medium- to high-sulfur
(2% to 3%) eastern coal plus occasional firing of low-sulfur waste
oil results in flue gas generally containing 500 ppm to 1,300 ppm
S02 by volume. Boiler No. 1, with a capacity of 45,000 kg/hr
(100,000 Ib/hr) of steam, was tested according to the scenario
shown in Table 1, page 4.
Scrubber System
Each boiler is equipped with a regenerable liquor, double-alkali
scrubbing unit. Table 10 lists design, operating, and performance
characteristics of the scrubbers.
25
-------
.,.- REMAINING -Jf'
^, '- PARTI CULATE>LUME
STEAM PLUME
COAL WEIGHING
DEVICE
27 - 38°C
(80 - 100°F)
SCRUBBER INLET
SAMPLING SITE
SCRUBBER EXHAUST
SAMPLING SITE
SPREADER
STOKER BOILER
WITH
ECONOMIZER
DUAL ALKALI
SCRUBBER
Figure 4. Schematic diagram of Boiler No. 1 and associated equipment
at General Motors Chevrolet Plant in Parma, Ohio.
-------
TABLE 10. DESIGN, OPERATING, AND PERFORMANCE CHARACTERISTICS
OF DOUBLE-ALKALI SCRUBBING SYSTEM
Application
Four coal-fired stoker boilers
Fuel characteristics
System design
Process mode
Pressure drop
Status
Startup date
Inlet gas conditions
Flow rate
S02
02
Particulate
S02 outlet concentration
S02 removal efficiency
Scrubbing solution characteristics
PH
Total sodium
Active sodium
Calcium ion
Soda ash makeup
Lime utilization
Filter cake solids
Filter cake disposal
Coal: 25,600 to 31,400 J/g (11,000 to
13,500 Btu/lb), 1.5% to 3.0% sulfur
(waste oil also burned, but in small
quantities compared with coal)
Four three-stage multiventuri flexitray
scrubber modules
Dilute active alkali
25 to 33 cm H20 (10 to 13 in. H20)
Operational
March 1974
30.9 ms/s at 27°C (65,500 acfm at 80°F)
800 to 1,300 ppm
Not available
0.7 g/m3 (0.3 gr/scf) (dry)
20 to 130 ppm
90% to 99%
5.5 to 7.5
0.58 to 0.96 molar
0.087 to 0.13 molar
305 to 490 ppm
>0.1 mole/mole S02 removed
1.32 to 1.90 mole/mole S in cake
40% to 55%
Offsite landfill
Nonsteady-state operations result in makeup rates of 0.028 to 0.05 mole
N02 per mole S02 removed.
27
-------
The primary function of the scrubber is to reduce S02 emissions in
the boiler flue gas by reaction with an alkali scrubbing liquor.
A secondary function is the simultaneous removal of particulate
matter. The functions of the liquor regeneration system are to
precipitate the reacted sulfur (in the form of sodium-sulfur com-
pounds) and vacuum filter the solids for disposal while simultane-
ously regenerating the scrubber liquor via an ion exchange
mechanism.
Sulfur dioxide is removed from the flue gas in the scrubber by
reaction with the alkali liquor, 0.1N NaOH. S02 is combined
chemically with the NaOH to form sodium bisulfite by the reaction
mechanism shown below.
2NaOH + S02 - *-Na2S03 + H20
Na2S03 + S02 + H20 - *-2NaHS03
Simultaneously some sodium sulfite reacts with the oxygen in the
flue gas to produce sodium sulfate:
Na2S03 + 1/202 - *» Na2S04
Scrubber liquor regeneration is achieved by reacting calcium
hydroxide (lime slurry) with the soluble sodium-sulfur compounds
formed in the scrubber. Calcium sulfite and calcium sulfate
precipitate out by the mechanism shown below while the sodium
remains in solution with the hydroxide, thus regenerating the
scrubber liquor.
2NaHS03 + Ca(OH)2 - ^Na2S03 + CaS03 + H20
Na2S03 + Ca(OH)2 - >-CaS03 1 + 2NaOH
Na2S04 + Ca(OH)2 - *-CaS04 | + 2NaOH
28
-------
Figure 5 shows a flow diagram of the scrubber system. Flue gases
enter through a prequench section at the bottom of each unit and
then flow in a countercurrent direction through an aqueous; sodium
hydroxide sulfite-bisulfite solution. The scrubber is a two-tray,
impingement-type unit with feed and recycle streams entering at
the top. It was modified from the original three-tray design in
an effort to reduce scaling. The absorption trays (Koch) have
movable self-adjusting bubble caps that respond to variations in
gas flow. Pressure drop through each unit is typically 19 cm
(7.5 in.) H20, and the maximum liquid-to-gas ratio is 2.7 L/m3
(20 gal/1,000 ft3). For control of entrained liquor, each unit
is equipped with a high-efficiency mist eliminator. Acting as
an impingement separator, the mist eliminator is composed of corru-
gated profile plates assembled with phase separating chambers.
The spent scrubbing liquor, containing absorbed sulfur compounds
and fly ash, is bled from the scrubber recycle tank to the chemical
mix tank where lime (CaO) is added for sodium hydroxide regenera-
tion. After a final clarification step to remove solids (calcium
sulfur compounds and fly ash) and a softening step to remove excess
calcium, the liquor is ready to be fed back to the scrubber. Vacuum
filters are used to dewater the sludge from the clarifier bottoms
after washing to remove NaOH.
MONITORING OF PLANT PROCESSES DURING TESTING
The boiler and scrubber processes were monitored during testing
for the following reasons:
1. To insure that the boiler was operating near capacity and at a
relatively steady state;
2. To insure that the boiler was operating within normal tolerances;
29
-------
NaOH
U)
o
FILTRATE
PUMP
VACUUM FILTERS
*»
CAUSTIC
STORAGE
SODA ASH
FEED PUMP
CARBON DIOXIDE
STORAGE TANK
Figure 5. Flow diagram of double-alkali scrubbing unit.
-------
3. To insure that the scrubber was operating normally at or near
the design pressure drop and that no liquor feed abnormalities
occurred;
4. To record process data used in determining emission rates
(pounds emitted/106 Btu); and
5. To record process data used as a qualitative guideline to
document normal boiler operation.
Eight process parameters were monitored during emissions testing.
Tables 11 through 14 are compilations of these data. All process
parameters recorded are described below.
Steam Production Rate
The steam production rate was monitored by two instruments on the
boiler control panel. The instantaneous values in Table 11 through
14 were recorded by a continuous monitor while the average values
at the bottom of the tables were obtained from a steam integrator
monitored over the entire test period. Fluctuating steam demand
on the power plant caused pressure variations in the steam lines
that slightly affect the accuracy of all steam production monitor-
ing equipment. The degree to which these fluctuations can affect
the accuracy is unknown.
Flue Gas Oxygen
The percent oxygen was sampled on a continuous basis at the flue
gas inlet to the economizer. The accuracy of this instrument was
considered poor by company personnel. It was monitored for unusual
fluctuations which might indicate abnormal boiler operation.
The flue gas oxygen meter in conjunction with the opacity meter
provided useful information for the boiler engineer. These
readings allowed the engineer to minimize smoke production by
31
-------
TABLE 11. POWER PLANT OPERATING PARAMETERS (11/12/80, RUN SI)
Time
5:34
5:35
5:50
6:12
6:24
7:02
7:19
7:34
7:50
8:00
8:02
Average
Average
Flue
Steam gas
production, oxygen,
Ib/hr %
(START)
93,500 5.2
89,500 5.7
93,000 4.9
Test interrupted in
92,500 5.2
95,000 4.9
92,000 5.1
92,000 5.0
92,500 5.0
(FINISH)
coal feed
steam production
Mechanical
Scrubber Inlet collector
pressure scrubber gas pressure
drop, Scrubber temperature, drop,
in. H20 pH °F in. H20
7.5
7.4
7.6
6.8 239 3.4
6.8 239 2.6
6.8 239 2.6
Gas Smoke
temperature, density,
°F %
502 2
500 1
504 1
order to change sampling train filters.
7.5
7.6
7.6
7.7
7.2
7,
90,
6.8 241 2.7
6.7 243 2.9
6.6 242 2.7
6.6 243 2.7
6.6 243 .2.8
415 Ib/hr
385 Ib/hr
509 1
510 1
505 1
508 1
508 1
-------
TABLE 12. POWER PLANT OPERATING PARAMETERS (11/13/80, RUN 1)
LO
Time
9:22
9:22
9:37
9:54
10:09
10:35
10:50
11:14
11:38
11:35
12:02
12:23
12:35
12:36
Average
Average
Flue
Steam gas
production, oxygen,
Ib/hr %
(START)
98,500
93,000
96,000
89,000
94,000
92,000
93,500
95,000
Outlet filter
94,000
94,000
99,000
(FINISH)
coal feed
steam production
5
5
5
5
5
5
5
5
.3
.2
.4
.1
.5
.4
.0
.1
change
5
5
5
.7
.5
.1
Scrubber
pressure
drop,
in. H20
8.2
8.2
8.3
8.1
8.2
7.5
7.4
7.5
reguired
7.7
7.8
7.7
7,500
92,850
Scrubber
PH
6.
6.
6.
6.
6.
6.
6.
6.
due to
6.
6.
6.
Ib/hr
Ib/hr
8
8
8
8
8
8
8
8
high
8
8
8
Inlet
scrubber gas
temperature,
op
255
255
255
253
254
253
251
253
loading.
256
260
263
Mechanical
collector
pressure
drop,
in. H20
3
3
3
3
3
3
3
3
3
3
3
.2
.2
.2
.2
.1
.2
.1
.2
.3
.4
.3
Gas
temperature ,
op
519
515
512
512
510
502
510
510
520
520
525
Smoke
density,
o,
o
3
2
1
1
2
1
1
8
4
5
25
-------
TABLE 13. POWER PLANT OPERATING PARAMETERS (11/13/80, RUN 2)
Time
6:51
6:52
7:09
8:10
8:31
8:55
9:15
10:26
10:57
11:15
11:15
Average
Average
Steam
production,
Ib/hr
(START)
93,000
93,500
95,500
94,000
95,000
97,000
95,000
94,500
96,000
(FINISH)
coal feed
Flue
gas
oxygen ,
5.7
5.3
5.6
6.0
5.3
5.8
5.8
5.6
5.7
steam production
Scrubber
pressure
drop,
in. H20
8.2
8.2
8.1
7.8
7.9
8.2
7.6
7.7
7.6
8,196
95,524
Scrubber
PH
6.9
7.0
7.1
6.6
6.8
6.8
7.2
7.8
7.4
Ib/hr
Ib/hr
Inlet
scrubber gas
temperature,
°F
264
270
260
265
265
270
268
270
270
Mechanical
collector
pressure
drop,
in. H20
3.5
3.5
3.6
3.4
3.5
3.5
3.6
3.6
3.6
Gas
temperature,
520
532
530
523
521
560
525
530
530
Smoke
density.
3
2
2
4
4
2
6
8
2
-------
TABLE 14. POWER PLANT OPERATING PARAMETERS (11/14/80, RUN 3R)
Time
7:08
7:08
7:28
7:55
8:21
8:57
9:26
9:47
10:14
10:14
Average
Average
Steam
production,
Ib/hr
(START)
87,000
92,000
93,500
94,000
96,000
94,000
93,000
91,500
(FINISH)
coal feed
Flue
gas
oxygen,
5.8
5.5
5.4
5.0
4.8
5.2
5.3
5.2
steam production
Scrubber
pressure
drop,
in. H20
7.4
7.4
7.2
7.2
7.3
7.4
7.3
7.3
7,677
93,065
Scrubber
PH
6.8
6.8
6.8
6.8
6.8
6.8
6.8
6.9
Ib/hr
Ib/hr
Inlet
scrubber gas
temperature,
Op
248
246
247
247
249
250
249
248
Mechanical
collector
pressure
drop,
in. H20
2.9
3.0
2.9
3.1
3.0
3.0
3.0 .
3.0
Gas
temperature ,
op
508
508
509
510
510
510
507
505
Smoke
density,
2
1
1
1
1
1
1
2
-------
adjusting the excess air flow to the boiler (as indicated by the
oxygen meter) until the optimum flow was achieved (as indicated
by a subsequent decrease in opacity). The oxygen meter also
indicated that if opacity increased without a corresponding
decrease in flue gas oxygen, other problems such as poor coal
feed stock, improper grate speed, or a combination of both could
be responsible for the increase in opacity.
Scrubber Pressure Drop
The scrubber pressure drop was measured across the inlet and outlet
on a continuous basis. These readings served as an indication of
plugging on the trays. An increase in pressure drop would indicate
that plugging has occurred in the scrubber. It should be noted
that the pressure drop will vary slightly with boiler load. If
plugging is thought to have occurred, comparisons must be made
with pressure drops recorded when the scrubber was known to be
unrestricted and operating under the same boiler load conditions.
Scrubber pH
The pH was monitored at the top tray of the two-tray countercurrent
flow scrubber. The pH was used as a primary means of controlling
the entire SO2 scrubbing system by controlling the amount of
scrubber feed liquor added to the recycle line. In turn, the
total volume of this flow determines how much bleed liquor is
sent to the chemical mix tanks, how much lime is added for re-
generation, how much sludge is formed in the clarifier, and how
much soda ash is needed for softening.
The original scrubber design called for a pH of 6.0 at the bottom
tray. Modifications to the original design, discussed in the
"Scrubber System" section, resulted in monitoring the top tray
pH and maintaining it at a level of 6.8.
36
-------
Scrubber Inlet Gas Temperature
The inlet gas temperature is monitored at the gas prequench section
of the scrubber. A high inlet temperature would indicate that the
quench section is not operating normally, resulting in a flue
gas temperature to the scrubber that is not within the normal
design range.
Mechanical Collector Pressure Drop
The pressure drop across the multiclone collector was monitored to
insure that it operated normally. A high pressure drop would
indicate that flow through the collector was restricted, possibly
due to obstructions in the cyclones or to hopper overloading.
It should be noted that boiler load affects the pressure drop
across the collector. When plugging was thought to have occurred,
the procedure described under "Scrubber Pressure Drop" was followed.
Gas Temperature
The flue gas temperature was monitored at the boiler outlet (inlet
to the economizer) on a continuous basis. The flue gas temperature
typically varies with boiler load.
Smoke Density
The smoke density was monitored at the mechanical collector outlet
by an outdated in-stack Baily opacity meter. Plant personnel
commented that the instrument's accuracy and reliability were
marginal. For this reason these data should not be used as a
quantitative measure of the unscrubbed flue gas smoke density.
However, the data were useful as a qualitative gauge, indicating
how well the boiler responded to adjustments made in an effort to
reduce smoke production.
37
-------
Average Coal and Steam Values
The average coal feed rates and steam production rates reported at
the bottom of Tables 11 through 14 were measured by integrators.
The steam integrator was located on the boiler control panel and
was subject to the same pressure fluctuations discussed under
"Steam Production Rate." The coal integrator was located on the
boiler feed hopper. This integrator measured how many batch
loads of coal [in 91-kg (200-lb) increments] were fed into the
boiler over an entire test period.
COMMENTS ON TESTS
Boiler Steam Production Rate
Throughout all of the tests, boiler load was maintained at a
relatively constant value. Although demand on the power plant as
a whole was fluctuating with time, steam demand on Boiler No. 1
was held steady by creating a false load (venting steam to the
atmosphere) when demand was low. This procedure allowed boiler
engineers to maintain a steady steam demand above 41,000 kg/hr
(90,000 Ib/hr; 90% capacity) for Boiler No. 1.
Coal Feedstock
Plant personnel reported that higher than normal amounts of coal
fines were in the coal feedstock during Runs 2 and 3. As a result,
smoke production was increased and adjustments normally used to
control smoke (underfire air, overfire air, and bed adjustments)
were ineffective. Observation of the coal bed revealed that a
significant portion of the bed, consisting of these coal fines,
would blow-off and become entrained in the flue gases to the
scrubber. Smoke density readings, as monitored by the in-stack
opacity meter, ranged between 3% and 8% throughout test No. 2. It
38
-------
should be noted that momentarily spikes up to as high as 25%
occurred during the final minutes of the test. The fines became
so high during Run 3 that the run was terminated prior to com-
pleting the traverse.
39
-------
SECTION 4
LOCATION OF SAMPLING POINTS
Two locations along the exhaust path of Boiler No. 1 were tested - a
site between the multiclones and the scrubber (the scrubber inlet)
and a site on the stack after the scrubber (the scrubber exhaust).
In addition, the coal being fed into the boiler was sampled. The
sampling sites are shown in Figure 4 (Page 26).
SCRUBBER INLET
Diagrams of this site are shown in Figure 6. It is located on a
1.8 mx 1.2 m (6 ft x 4 ft) ID rectangular duct located below a
damper and elbow coming from the multiclones and the boiler. Three
7.6 cm (3 in.) ID sampling ports are located on this duct, about
2.7 m (9 ft) below the damper and about 0.8 m (2 1/2 feet) above
a widening of the ductwork leading to ID fans and the scrubber.
The ports are about two equivalent duct diameters downstream of a
flow disturbance and half a diameter upstream. A 49-point traverse
in a 7 x 7 matrix would be needed for Method 5B testing by the full
EPA reference method traverse procedures. Such a traverse would
make the installation of additional sampling ports necessary. Be-
cause EPA needed results quickly, a decision was made to sample
sixteen points along each of three diameters in line with the
existing ports, instead of installing new ports for the Method 5B
samples.
The Andersen particle size samples were taken through the center
port, 51 cm (20.1 in.) from the inside wall of the duct. This
41
-------
SIDE VIEW
DETAIL A
1/2" PIPE
DAMPER (DISTURBANCE)
1/2 PIPE (PRESSURETAP)
WHICH MAY INTERFERE
GRATING
TO SCRUBBER
3"PORTS. TYPICAL FEMALE
THREADS ~ 3" DEEP WITH
MALE PLUG TO SEAL
1/2" PI PECAN BE
REMOVED
-4'
DETAIL "A
SECTION A-A
TOP VIEW
NOTE:-8ft. IN FRONT OF PORTS TO
ANOTHER DUCT
NOTES: (1)~ 4" INSULATION ON ALL SIDES OF DUCT
(2) 3rd LEVEL, SAMPLE LOCATION INSIDE BUILDING, POWER
~ 50 FEET AWAY (SPECIAL ADAPTERS WILL BE FURNISHED)
(3) GOOD AREA FOR CLEAN-UP
Figure 6. Scrubber inlet sampling site at Parma, Ohio GM Plant.
-------
location has a velocity approximately equal to the arithmetic mean
of the velocities at the 48 traverse points.
SCRUBBER EXHAUST
Figure 7 provides a side view and a top view of the site on the
roof of the boiler house, where scrubber exhaust was sampled. Three,
7.6 cm (3 in. ID) sampling ports were located on three sides of a
1.3 m (53 in. ID) stack at this site. These ports were about seven
stack diameters above a fitting connecting the scrubber to the stack
and about four stack diameters below the mouth of the stack.
During full Method 5 and 5B traverses, 12 points were sampled along
each of two perpendicular diameters of the stack.
The Andersen particle size sample was taken at a location having a
velocity approximately equal to the arithmetic mean of the veloc-
ities measured at the 24 Method 5 and 5B traverse points. This
location is 47.2 cm (18 5/8 in.) from the inside wall of the stack,
from port B (shown in Figure 7).
For the Method 5 and 5B sampling at a single point, the nozzles
were initially placed next to each other at the center of the stack.
Before the sampling began, the probes and nozzles were then moved
back far enough apart so that there was no evidence (from the pitot
tube pressure drops) of either train interfering with the other
train's pitot tube velocity measurements. The nozzles are estimated
to have been within 7.6 cm (3 in.) of each other during the single
point tests.
COAL SAMPLING SITE
Coal was sampled with a shovel from the middle port of the coal
feed chute leading directly into the boiler coal feeder overthrow
rotor. This site is located after the coal weighing device.
43
-------
N
-H 1—
LADDER
SHED
WITH TABLE
DISTURBANCE
18'
ROOF
PORT B
PORT A
H h
LADDER
•4'-6"
O
3" PORTS
(COUPLINGS. 2.5"
TO INSIDE WALL)
31'
\
*- DISTURBANCE
TOP VIEW
SIDE VIEW
NOTE: TWO CONSOLES CAN FIT INSIDE. POWER
70ft. AWAY -100ft. TO GROUND.
EQUIPMENT MAY BE BROUGHT UP BY ROPE.
Figure T. Scrubber exhaust sampling site at Parma, Ohio GM Plant.
-------
SECTION 5
DESCRIPTION OF SAMPLING TRAINS
PARTICULATE SAMPLING TRAINS
EPA Method 5 sampling trains were used for determining Method 5
and Method 5B particulate mass emission rates (see Figure 8).
The temperature control system settings on the trains used for
Method 5B sampling were set to maintain filter temperatures of
177 ±14°C (350 ±25°F) instead of the 120 ±14°C (298 ±25°C) used
for Method 5 sampling. Heated glass-lined probes and Reeve Angel
Type 934 AH filters were used for the Method 5 and 5B testing.
The filter temperatures were monitored using thermocouples in-
stalled on the back halves of the filter holders of each sampling
train. A heated glass cyclone was used between the sampling
probe and the filter holder to minimize filter changes.
PARTICLE SIZE DISTRIBUTION SAMPLING APPARATUS
Sampling for particle size was performed using an Andersen cas-
cade impactor with seven stages and a back-up filter. Figure 9
is a schematic drawing of the sampling train. It consisted of
the following equipment, listed in order of the air flow: a
0.47 cm (0.187 in.) diameter probe tip; a standard Andersen head;
a 1.2 m (4 ft) stainless steel probe; a Smith-Greenburg impinger
with water, then an impinger charged with color-indicating silica
gel; an EPA-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 Andersen
sampler.
45
-------
.TEMPERATURE
SENSOR
0 75 TO 1 in
THERMOMETER
CHECK VALVE
THERMOMETERS
VACUUM LINE
ACUUM GUAGE
HEATED AREA
FILTER HOLDER
NOZZl E/HEATED -
PROBE £
REVERSE TYPE
PI TOT TUBE
PITOT MANOMETER
IMPINGERS
ORIFICE
AND
MANOMETER
ICE BATH
Figure 8. Particulate sampling train - EPA Methods 5 and 5B.
-------
Cascade
Impaclor
Nozzle •
Stack Wall
,
L
Probe
t
Reverse-Type
Pilol Tube
T^X
Thermometer
Vacuum
Tubing
Check Valve
Ice
Water
Ull
~J f
.Silica
Gel
Pilot Manometer/ Thermometers BVPass Valve Impingers
Orifice.
Vacuum Gauge
Air Tight Pump
Vacuum
Line
Figure 9. Particle size distribution sampling apparatus - Andersen 2000, Inc.
-------
SECTION 6
SAMPLING AND ANALYTICAL PROCEDURES
SUMMARY
The general sampling and analytical procedures used and the number
of tests performed are described in Section 2 of this report.
Particulate emissions downstream of the scrubber 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 content of the
particulates collected was measured by techniques supplied to MRC
by EPA October 10, 1980. The plume opacity was measured by EPA
Method 9. Grab samples of coal were taken before, during, and
after every Method 5 and 5B emission test and analyzed by ASTM D3176
ultimate analysis, ASTM D2015-66(72) heating value analysis, and a
modified version of ASTM D410-38 sieve size analysis. Further
details on the sampling, analysis, and quality assurance procedures
are shown below.
FIELD SAMPLING
The Method 5 and 5B testing during each run was approximately
simultaneous. A decision was made by EPA not to shut down all
sampling trains when filters were changed on one of the trains.
This resulted in small deviations from true simultaneous sampling.
The exact times that samples were taken, filters changed, etc. are
provided in the field data sheets in Appendix A.4.
49
-------
Each Andersen sampler run was conducted for 5-8 minutes under iso-
kinetic conditions. After the completion of each run, the Andersen
heads were opened and the substrates inspected and stored for weigh-
ing at MRC's laboratory.
Coal samples were taken with a shovel through the middle sampling
port and stored in 1.9 x 10~3 m3 (1/2 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 distilled
water and bottled. The acetone rinses were bottled separately from
the water rinses. Access to the samples was limited in the field by
storing them in locked MRC vehicles, except when they were being
handled by authorized individuals. The samples were shipped to
the MRC Dayton Laboratory for analysis in the vehicles used for
storing them in the field. MRC's usual security procedures were
used to limit access to the samples while they were being analyzed.
Records of the chain-of-custody of the samples were maintained.
CHEMICAL ANALYSIS
The detailed procedure used for analyzing the sulfate and sulfuric
acid content of the Method 5 and 5B stack samples are summarized
in Appendix E, along with the results of tests of the procedure.
The procedure used is based on the procedure supplied to MRC by
EPA at a meeting 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; and the filter in isopropanol was
50
-------
ultrasonically extracted for 30 minutes. The extracts were filter-
ed and analyzed for sulfuric acid, using the barium-thoriri titration.
This isopropanol extraction and analysis procedure was repeated twice
on every sample. Those portions of the extracts that were not
titrated were bottled and retained.
After the sulfuric acid extraction, water was added to the filter
and solid residue; they soaked for at least 12 hours; and were
ultrasonically extracted for 30 minutes. The extract was filtered,
passed through a Rexyn 101 ion exchange column, and analyzed for
sulfates (using barium-thorin titration). Four Method 5 or 5B
filters were extracted twice with water and analyzed twice for
sulfates, as an indicator of the effectiveness of the extraction
procedure. The remaining water extracts were also retained.
The solid residues from these extractions were dried and saved under
dry nitrogen in a refrigerator, along with the untitrated extracts.
These samples were saved until May 12, 1981.
Blank filters, residue from the evaporation of clean acetone and
residue from the evaporation of clean water (the solvents used for
probe washing) were also analyzed when the stack sample 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.
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.
51
-------
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 material,
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 lb/106 Btu emissions were taken from the boiler
emission regulations in 40 CFR 60.45(f)(4)(ii).
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 and the
results of tests of the methods.
Appendix F contains the detailed analytical data.
Appendix G identifies the people performing the sampling, analysis,
and data reduction.
52
-------
QUALITY ASSURANCE
Quality assurance and control measures 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.
Particle deposits on the stages of the Andersen Sampler were
checked to assure that the stages remained tight during sampling.
A certified smoke inspector was used for opacity monitoring by
EPA Method 9. Standard ASTM procedures were used for the coal
analysis.
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 by the MRC Laboratory
under the supervision of the MRC QA/QC Supervisor.
Three sulfate filters and one sulfuric acid audit sample were run
blind by the analysts prior to the receipt of the samples from
Parma to test the procedures used, to study recovery efficiencies,
and to study the precision of the sulfate analysis method at the
MRC Laboratory.
Additional audit samples as well as blank filters, clean acetone
samples, clean water samples, and calibration standards were run
along with the Parma stack samples. All blanks and quality assur-
ance audit samples were handled in the same manner as the field
samples.
Standard ASTM procedures were used for the coal analysis. To
obtain an indication of the precision of the coal analysis data,
half of the raw coal sample taken during the second Andersen run
at the inlet was relabeled as a sample from Parkersburg, West
Virginia and analyzed along with the coal samples for an emission
53
-------
test in Parkersburg. Because it was a leftover half sample, the
sample was not split in half prior to the second analysis. This
was the only change from the procedure used for the actual
Parkersburg samples. The chemists analyzing the coal 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 the first 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 reproducibility 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) sug-
gested that the small sizes of the samples and real differences
in the composition of coal from one piece to another were the main
causes of the poor reproducibility of the results of the analysis
of different "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.
54
------- |