United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
Research Triangle Park NC 27711
EPA-600/4-85-004
January 1985
Research and Development
vvEPA
A Summary of the
1983 EPA National
Performance Audit
Program on Source
Measurements
-------
EPA-600/4-85-004
January 1985
A SUMMARY OF THE 1983 EPA NATIONAL PERFORMANCE AUDIT PROGRAM
ON SOURCE MEASUREMENTS
by
E.W. Streib, T.J. Logan and M.R. Midgett
Quality Assurance Division
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
FOREWORD
Measurement and monitoring research efforts are designed to anticipate
potential environmental problems, to support regulatory actions by develop-
ing an in-depth understanding of the nature and processes that impact health
and the environment, to provide innovative means of monitoring compliance
with regulations, and to evaluate the effectiveness of health and environ-
mental protection regulations through the monitoring of long-term trends.
The Environmental Monitoring Systems Laboratory, Research Triangle Park,
North Carolina, has responsibility for: assessment of environmental moni-
toring technology and systems; implementation of agency-wide quality assur-
ance programs for air pollution measurement systems; and supplying technical
support to other groups in the Environmental Protection Agency, including
the Office of Air and Radiation, and the Office of Toxic Substances.
The major task of this study was to report the results of the national
quality assurance audit program for stationary source test methods. Audits
were designed to estimate the minimal analytical and computational accuracy
that can be expected with EPA Method 5 (dry gas meter only), Method 6 (sul-
fur dioxide), Method 7 (nitrogen oxides), Method 19 (coal), and Method 3
(carbon dioxide and oxygen). Statistical analysis was used to characterize
the data.
Thomas R. Hauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina
iii
-------
ABSTRACT
In the spring and fall of 1983, the Quality Assurance Division con-
ducted the National Audits for Stationary Source Test Methods. The audit
materials consisted of: a calibrated orifice for Method 5 (dry gas meter
only), five simulated liquid samples each for Method 6 (SC>2) and Method 7
(NOX), two coal samples for Method 19A, and a disposable gas cylinder for
Method 3 (Orsat analyzer). Participating laboratories sent their data to
the Source Branch and in return received a written report comparing their
results to EPA's.
In the Method 5 spring audit, the mean value for all participants
differed by 5.7% from the true (EPA) value. For the fall audit, the par-
ticipants' mean was 4.1% from the true value. In the two Method 6 audits,
the average mean differed by 3% from the true value. The average mean in
the two Method 7 audits was 15% of the true value.
In the two coal audits, the parameters measured were sulfur, moisture,
ash, and Btu. On the average for the sulfur analysis, 85% of the partici-
pants measured within 10% of the expected value; for Btu, 100% of the parti-
cipants measured within 10% of the expected value.
In each of the two Method 3 audits, each parameter had only one con-
centration. The means for C02 were within 10% (spring) and 6% (fall) of
the expected value. For both audits, the mean for 02 was less than 1%.
This report covers the period from January 1983 to December 1983.
iv
-------
CONTENTS
Page
Foreword iii
Abstract iv
Figures vi
Tables vii
Acknowledgments viii
1. Introduction 1
2. Summary 2
3. Method 5 Dry Gas Meter Audit 6
4. Method 6 Audit 12
5. Method 7 Audit 14
6. Method 19A Coal Audit 16
7. Method 3 Audit 20
References 22
Appendices
A. Frequency Distributions 23
B. Instructions for EPA Audit Materials 31
-------
FIGURES
Number Page
1 Cumulative accuracy for participants in the
Method 5 audits, 0383 and 0983 8
2 Previous results of Method 5 audit ..... 9
3 Results of the Method 5 audit 0383 10
4 Results of the Method 5 audit 0983 11
vi
-------
TABLES
Number
1 Participants' Results for Method 5 Audit
(All Data - No Outliers Removed) 2
2 Participants' Results for Method 6 Audit
(All Data - No Outliers Removed) 3
3 Participants' Results for Method 7 Audit
(All Data - No Outliers Removed) 3
4 Participants' Results for Method 19A Coal Audit
(All Data - No Outliers Removed) 4
5 Participants' Results for Method 3 Audit
(All Data - No Outliers Removed) 5
6 Method 5 Audit Participants 6
7 Method 6 Audit Participants 12
8 Summary of Source SC>2 Audits 13
9 Method 7 Audit Participants . 14
10 Summary of Source NOX Audits 15
11 Coal Audit Participants 17
12 Source Coal Audit - 0383 18
13 Source Coal Audit - 0983 19
14 Method 3 Audit Participants 20
15 Source Method 3 Audit - 0583 21
16 Source Method 3 Audit - 1183 21
vii
-------
ACKNOWLEDGMENTS
We express our appreciation to the laboratories that participated in
our audits. Thanks also to the Standards Laboratory (EPA/EMSL/ Performance
Evaluation Branch) who did our Acceptance Testing on the audit samples.
For providing the data systems to store and evaluate the data, we thank the
programmers of the Data Management and Analysis Division/EMSL.
viii
-------
SECTION 1
INTRODUCTION
The Environmental Monitoring Systems Laboratory (EMSL) of EPA, in 1977,
established a performance audit program to evaluate the performance of com-
panies that conduct compliance testing, using EPA Reference Methods. The
audits verify the analytical accuracy of EPA Reference Methods 3, 6, 7, and
19A and the calibration accuracy of the Method 5 control console (1).
Accuracy is defined as the percent difference between the participant's
analytical results and the EPA expected value. By participating in this
(free and voluntary) program, testing companies can compare their perform-
ance to other laboratories conducting similar measurements.
Two audits were conducted for each of Methods 3, 5, 6, 7 and 19A in
1983. Every participating laboratory received an audit package consisting
of the audit sample, a data card, instructions, and an envelope for return-
ing the data to EPA. A label for returning the audit device was also
included for the Method 5 audits. Participants had 8 weeks to return data to
EPA. At the end of this period all data received were statistically ana-
lyzed to determine the precision and accuracy obtained by the participants.
Participants were given 8 weeks to return the data, which was then statis-
tically analyzed for precision (repeatability) and accuracy.
The Quality Assurance Division of EMSL maintains a repository of audit
samples for the EPA Methods 3, 6, 7 and for coal. These (stable) samples
are available to any laboratory needing them (for example) to train new
personnel or to conduct quality control checks in the laboratory. Since
the expected values (for these samples) are included with the analysis in-
structions, there is no requirement for the data to be returned to the EPA.
However, we recommend that participants make use of this sample repository
to help improve their overall analytical skills.
This report summarizes the results obtained in the 1983 source audits.
-------
SECTION 2
SUMMARY
In the spring and fall of 1983, EPA's EMSL, Research Triangle Park,
North Carolina, conducted National Quality Assurance Audits for Stationary
Source Test Methods 5 (dry gas meter only), 6 (S02>, 7 (NOX), 19A (coal),
and 3 (Orsat analyzer). Industrial laboratories, contractors, foreign lab-
oratories, as well as local, state, and federal agencies participated.
Two Method 5 audits were conducted in 1983. The overall results (no
outliers removed) are summarized in Table 1. In the first audit, the mean
for all participants was 5.7% from the true value and in the second audit
it was 4.1%. After correcting for outliers, the means for the 0383 and 0983
audits were 2.6% and 2.IX, respectively, from the true value. The partici-
pants' performances were not statistically different from previous audits
(2,3,4).
TABLE 1. PARTICIPANTS' RESULTS FOR METHOD 5 AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of
sample
Orifice
Audit
Parameter date
Volume 0383
0983
No. of
analyses
763
614
Mean
(% from
5.7
4.1
Median
EPA value)
2.2
1.9
Std.
dev.
32.7
21.3
Table 2 presents the data (no outliers removed) from the two 1983
Method 6 audits. This audit procedure requires the participants to deter-
mine the sulfate content in five aqueous solutions using the titration pro-
cedure of Method 6. For each concentration in the 0283 audit, the mean of
the participants was 4% higher than the true value, in contrast, the median
differed by less than 1%. In the 0883 audit, the means differed by 3%, and
the medians again were less than 1%. In both audits, 50% to 68% of the
participants achieved an accuracy within 2% for 9 out of 10 of the samples.
However, on one low concentration sample, only 43% achieved this accuracy.
-------
TABLE 2. PARTICIPANTS' RESULTS FOR METHOD 6 AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of Audit
sample date
Aqueous 0283
sulf ate
0883
No. of
analyses
113
113
113
112
113
88
90
90
90
90
EPA
true
value
343.20
762.60
1143.90
1906.50
2A78.50
396.60
594.80
633.00
1349.80
1426.10
Participant results
Mean
361.54
794.51
1188.32
1971.14
2579.04
393.57
589.35
621.39
1314.46
1384.05
Median
342.80
758.30
1140.00
1890.00
2456.00
395.50
595.15
629.35
1339.45
1416.65
Std.
dev.
109.11
233.13
349.57
580.91
778.94
50.93
67.21
89.04
168.73
192.16
Table 3 presents the data (no outliers removed) from the two Method 7
audits in 1983. This audit procedure requires that the participants deter-
mine the nitrate content in five aqueous solutions. For each concentration
in the 0483 audit, the mean of the participants was 22% higher than the true
value, but in contrast, the median differed by less than 3%. In the 1083
audit, the means were 12% higher than the true value, and again the median
value differed by only 3%. In both audits, 34% to 38% of the participants
achieved an accuracy within 3% of the true value for 8 out of 10 samples.
However, on the two lowest concentration samples only 25% achieved this
accuracy.
TABLE 3. PARTICIPANTS' RESULTS FOR METHOD 7 AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of Audit
sample date
Aqueous 0483
sulfate
1083
No. of
analyses
59
62
63
62
63
53
53
54
54
52
EPA
true
value
59.70
298.70
477.80
737.20
955.60
118.40
218.40
240.30
407.20
454.80
Participant results
Mean
71.66
348.99
581.76
892.88
1159.22
128.12
235.98
267.01
440.13
511.48
Median
58.10
302.80
480.00
750.00
975.60
121.00
221.40
245.70
407.45
469.65
Std.
dev.
68.67
280.20
579.80
720.34
1005.48
32.76
64.95
71.78
136.98
143.74
-------
Table 4 summarizes the results of the two coal audits. Participants
analyzed each coal sample in duplicate for percentage of sulfur, moisture,
and ash, and for gross calorific value (Btu/lb). The means of the ash,
moisture, and sulfur content were within 4% of the expected value on both
concentrations. An accuracy of IX was achieved on the Btu content.
TABLE 4. PARTICIPANTS' RESULTS FOR METHOD 19A COAL AUDIT
(ALL DATA - NO OUTLIERS REMOVED)
Type
of Audit
sample date
No. of EPA true
Parameter analyses value
Participants' Results
Std.
Mean Median dev.
Coal 0383 % S
0983
0383
0983
0383 Z H20
0983
0383
0983
0383 % Ash
0983
0383
0983
0383 Btu/lb
0983
0383
0983
125
112
127
113
125
113
127
114
123
112
125
113
125
112
123
111
0.74
0.74
2.98
2.95
2.05
1.62
6.13
5.75
7.40
7.31
11.42
11.33
12719.00
12470.00
14165.00
14059.00
0.74
0.75
2.89
2.85
2.06
1.59
5.97
5.86
7.35
7.60
11.37
11.61
12643.50
12574.22
14115.97
14068.90
0.74
0.73
2.91
2.87
2.05
1.68
6.16
6.10
7.33
7.29
11.42
11.40
12684.00
12597.50
14141.00
14093.00
0.05
0.21
0.17
0.21
0.71
0.37
0.53
0.71
0.41
3.35
0.24
2.59
251.79
207.38
150.59
170.01
The results of the Method 3 audits are summarized in Table 5. Partici-
pants analyzed the gas sample twice for percentage of carbon dioxide,
oxygen, and carbon monoxide. The mean value of carbon dioxide in the 0583
audit differed by 11% from the expected value; whereas, in 1183 the mean
value was within 6%. In both audits, the mean values for oxygen were
within 1%. In contrast, the mean values for carbon monoxide differed as much
as 21% from the expected value.
-------
TABLE 5. PARTICIPANTS' RESULTS FOR METHOD 3 AUDIT
(ALL DATA - NO OUTLIERS REMOVED)
Type of Audit No. of Repli-
sample date Parameter analyses cate
EPA Participants' results
true Std.
value Mean Median dev.
Small 0583
cylinder
Z C02
Z 02
Z CO
1183 Z C02
Z 02
Z CO
64
63
64
63
53
53
52
51
52
51
40
40
1
2
1
2
1
2
1
2
1
2
1
2
5.00
5.00
15.10
15.10
1.00
1.00
7.10
7.10
11.20
11.20
0.50
0.50
4.49
4.59
15.12
15.10
0.81
0.79
6.72
6.70
11.23
11.20
0.47
0.47
4.70
4.70
15.05
15.00
0.80
0.90
6.80
6.80
11.40
11.40
0.50
0.40
0.73
0.98
1.22
1.22
0.30
0.33
0.34
0.53
1.20
1.27
0.23
0.31
-------
SECTION 3
METHOD 5 DRY GAS METER AUDIT
In the Method 5 audit procedure, participants use a calibrated orifice
to check the calibration of the dry gas meter in their EPA Method 5 control
console (meter box). They insert the orifice in the Method 5 meter box,
allow the box to warm-up, and then make three 15-minute volume measurements.
The participants convert each of the three volumes to cubic meters at
standard conditions using the formula specified in Method 5 (Eq. 5.1 of
Appendix A 40 CFR 60) and record them on the data card. Finally, they mail
the orifice and the data card to EPA for statistical analysis.
In the spring audit (0383), 71% of the 160 laboratories that received
the audit package returned data. In the fall audit (0983), 65X of the 153
laboratories returned data. These percentages are slightly lower than those
found in previous audits (2,3,4). Table 6 (which classifies the partici-
pants into general categories) compares the number of participants who re-
quested to participate in the Method 5 audit with the number who actually
returned data.
TABLE 6. METHOD 5 AUDIT PARTICIPANTS
Total
No. requesting samples
160
153
No. returning data
Contractor
University
Industry
Foreign
Federal
State
Local
0383
88
1
44
3
3
15
6
0983
80
0
43
4
3
17
6
0383
59
0
33
3
3
12
4
0983
43
0
32
4
2
15
4
114
100
Figure 1, a cumulative histogram, shows the accuracy obtained by par-
ticipants in the 0383 and 0983 Method 5 audits, expressed as the percent
difference from the true (EPA) value at various levels of accuracy. The
Code of Federal Regulations (1) requires that the dry gas meter be cali-
brated with an accuracy of ±2%. Figure 1 also shows that 46% of the
reporting laboratories in the 0383 audit and 53Z in the 0983 audit obtained
-------
this accuracy. These results indicate a slight increase to those in previ-
ous audits (Figure 2). Seventy-three of the laboratories participated in
both audits.
The histograms in Figures 3 and 4 compare the individual results of
the 0383 and 0983 audits with the mean and the median values for all parti-
cipants. The majority of the laboratories (0983) reported values higher
than the EPA value. The standard deviation of the triplicate analysis
(precision) by each laboratory indicated that for the 0383 audit, 65% of
the standard deviations for each set were within 0.3%. For the 0983 audit,
70% of the standard deviations were within 0.3%. Three percent of the 0383
and 4% of the 0983 data were identified as outliers using Chauvenet's
Criterion (5). Before the outliers were removed, the mean values for the
0383 and 0983 data differed by 5.7 and 4.1% from the true value, respec-
tively. After deletion of outliers, these values were reduced to 2.6 and
2.1, respectively.
-------
100
80
O i
SB70
§*
o 2
c« I
"J u
K O
O <
< °
z: UJ
K
O Sb
oe u
so
60
«»
30
Q.
w 20
10
OAUDIT0383.n = 763
EiAUDIT0983,n = 614
8
9
234 667
DIFFERENCE IN MEASUREMENTS FROM EPA VALUE, percent
10
Figure 1. Cumulative accuracy for participants in Method 5 audits 0383 and 0983.
-------
55
50
45
to
UJ
£
I
K
O
CD
40
35
30
I 1 I I
DEGREE OF ACCURACY = 2 PERCENT
I
I
I
I
I
I
0280/622 0880/725 0381/738 0981/723 0382/827 0982/769 0383/763 0983/614
AUDIT/NUMBER OF SAMPLES
Figure 2. Previous results of Method 5 audit.
-------
20
NUMBER OF VOLUMES = 763
MEDIAN = 1.4
15
8
co
IU
E 10
K
O
0
rt.
OUTLIERS
REMOVED
J.
JL
MEAN = 4.3
J_
OUTLIERS
REMOVED
>9 -8 -6 -4-2 0 2 4 6 8
DIFFERENCE FROM EPA VALUE, percent
Figure 3. Results of Method 5 audit 0383.
10 <10
10
-------
20
15
V)
IU
K
O
CD
2
10
NUMBER OF VOLUMES = 614
MEDIAN =-0.8
OUTLIERS
REMOVED
I
MEAN = 1.0
OUTLIERS
REMOVED
J 1 I
>9 -8 -6 -4 -2 0 2 4 6 <8
DIFFERENCE FROM EPA VALUE, percent
Figure 4. Results of Method 5 audit 0983.
11
-------
SECTION 4
METHOD 6 AUDIT
This audit checks the participant's ability to analyze Method 6 sam-
ples for sulfate. The audit set consists of five dilutions of 10 N sulfuric
acid in 25 ml sealed glass ampoules. The analyst withdraws 5.0 ml from
each ampoule, adds 30 ml of 3% hydrogen peroxide, and dilutes this sample
to 100 ml with distilled water. A 20 ml aliquot is then withdrawn from the
diluted sample, 80 ml of 100% isopropanol and thorin indicator are added,
and the sample is titrated with barium perchlorate (BalClO/,^) to a pink end
point. In calculating the results, the participants assume an original sam-
ple volume of 100 ml, and a sample of 0.021 dry standard cubic meter of
stack gas.
Table 7 (which classifies the participants into general categories)
compares the total number of participants requesting participation with the
number returning data. In the spring audit (0283), 71% of the 160 labora-
tories that received the audit package returned data. In the fall audit
(0883), 61% of the 147 laboratories returned data. Seventy-five labora-
tories participated in both audits and returned data.
TABLE 7. METHOD 6 AUDIT PARTICIPANTS
No. requesting samples No. returning data
0283 0883 0283 0883
Contractor
Industry
Foreign
Federal
State
Local
85
46
3
1
16
9
76
44
4
1
14
8
54
35
2
1
13
8
41
28
4
0
11
6
Total 160 147 113 90
Table 8 shows the percentage of laboratories that achieved 2% and 5%
accuracy for each of the five different concentrations in the 1983 Method 6
audits. It also shows that 55% of the reporting laboratories in the 0283
audit achieved an accuracy within 2% for the three higher concentrations.
However in the 0883 audit, 59% of the laboratories achieved an accuracy
12
-------
within 2X for all concentrations. Approximately 81% of the laboratories
were able to achieve an accuracy level of within 5% on 9 of the 10 samples.
TABLE 8. SUMMARY OF SOURCE S02 AUDITS
0283 0883
Concentration ± 2% ± 5% t 2% ± 5%
0 - 500 mg/DSCM 43% 73% 59% 82%
501 - 1000 mg/DSCM 50% 83% 68% 81%
68% 84%
1001 - 1500 mg/DSCM 58% 84% 61% 84%
60% 86%
1501 - 2000 mg/DSCM 55% 88%
2001 - 3000 mg/DSCM 55% 88%
n 113 90
13
-------
SECTION 5
METHOD 7 AUDIT
This audit checks the participant's ability to analyze Method 7 sample
for nitrate. The NOX audit set consists of five dilutions of a potassium
nitrate (KN03) solution in 25 ml glass ampoules which are autoclaved after
sealing so that bacteria which might attack the nitrate are destroyed. The
analyst withdraws 5.0 ml from an ampoule, adds this with 25 ml of the
Method 7 absorbing solution to a flask, adjusts the pH with sodium hydroxide,
and dilutes to 50 ml with distilled water. A 25 ml aliquot is withdrawn
from the diluted sample, placed in an evaporating dish, and treated as
described in Section 4.3 of Method 7. After this treatment is completed,
the absorbance is measured at 410 mn with a calibrated spectrophotometer.
In calculating the concentrations present, the participant assumes that
2000 ml of stack gas was sampled.
Table 9 shows both the total number of laboratories requesting parti-
cipation and the number that returned data for Method 7 audits 0483 and
1083. In the spring audit (0483), 51* of the 124 laboratories receiving
the audit package returned data. In the fall audit (1083), 49% of the
laboratories returned data. Forty-two laboratories participated in both
audits and returned data.
TABLE 9. METHOD 7 AUDIT PARTICIPANTS
No. requesting samples No. returning data
Contractor
Industry
Foreign
Federal
State
Local
0483
75
32
1
1
8
7
1083
67
31
2
1
5
5
0483
30
21
1
1
5
5
1083
22
21
2
1
4
4
Total 124 111 63 54
The percentage of laboratories that can achieve 5 and 10% accuracy for
each of the five concentrations are shown in Table 10. In addition, 32%
of the reporting laboratories in the 0483 audit achieved an accuracy within
5X on the lowest concentration. In the 1083 audit, 47% of the laboratories
14
-------
achieved an accuracy within 5%. Sixty-six percent of the laboratories were
able to achieve 10% accuracy on 9 of 10 samples in both audits.
TABLE 10. SUMMARY OF SOURCE NOV AUDITS
0483 1083
Concentration ± 5% ± 10% ± 5% ± 10%
0 - 200 mg/DSCM 32% 51% 47% 66%
201 - 400 mg/DSCM 52% 74% 58% 77%
63% 83%
401 - 600 mg/DSCM 46% 71% 63% 81%
42% 73%
601 - 800 mg/DSCM 52% 66%
801 - 1000 mg/DSCM 52% 68%
n 63 54
15
-------
SECTION 6
METHOD 19A COAL AUDIT
Subpart Da of 40 CFR 60 allows coal sampling and analysis to serve as
an acceptable method for demonstrating scrubber inlet flue gas sulfur
concentration. The coal audit checks the participant's ability to analyze
a coal sample for sulfur, ash, moisture, and BTU content.
Each set of coal samples consisted of two bottles, each containing 50
grams of 60-mesh coal. The participants analyzed each sample for sulfur,
moisture, ash, and gross calorific content. The following American Society
for Testing and Materials (ASTM) procedures (6) were recommended but not
necessarily mandated for participant's use in analyzing the coal samples:
ASTM D-3177 (Standard Test Method for Total Sulfur in the Analysis
of Coal and Coke);
ASTM D-3174 (Standard Test Method for Ash in the Analysis Sample of
Coal and Coke);
ASTM D-3173 (Test for Moisture in the Analysis Sample of Coal); and
e ASTM D-2015 (Standard Test Method for Gross Calorific Value of Solid
Fuel by the Adiabatic Bomb Method).
The participants measured the parameters and reported their results for
moisture (%) on an as-received basis, and their results for sulfur (%), ash
(%), and gross calorific value (Btu/lb) on a dry basis.
In the spring audit (0383), 85% of the 130 laboratories that received
the audit package returned data. In the fall audit (0983), 84% of the 117
laboratories returned data. Eighty-three laboratories participated in both
audits and returned data. Table 11 shows the total number of laboratories
requesting participation versus the number that returned data for coal
audits 0383 and 0983.
16
-------
TABLE 11. COAL AUDIT PARTICIPANTS
No. requesting samples No. returning data
Contractor
Industry
Federal
State
Local
0383
72
41
1
11
5
0983
62
40
0
11
4
0383
60
37
0
9
5
0983
51
34
0
9
4
Total 130 117 111 98
Tables 12 and 13 summarize the coal audit results. The number of anal-
yses is greater than the number of participants because some companies had
more than one laboratory participating. In this case, each laboratory
received its own set of samples and was asked to analyze the samples in
duplicate. Accuracies of 5% and 10% were chosen as the reporting criteria
for each of the four parameters (sulfur, moisture, ash, and gross calorific
content).
In both audits, 54% of the laboratories were able to analyze the sul-
fur content of the low level samples within 5% of the expected value.
Sixty-three percent achieved 5% accuracy for the high level sulfur concen-
trations. Only 28% achieved 5% on the low moisture concentration; however,
65% achieved 5% on the higher concentration in the 0383 audit. For the ash
analysis and Btu content, 92% to 99% of the reporting laboratories were
able to achieve an accuracy within 5% for both sample concentrations.
The participants' accuracy improved with higher concentrations on all
parameters except ash, where there was a slight decrease. For those labo-
ratories that reported duplicate analyses, the intra-laboratory precision
(repeatability) showed no correlation with concentration level.
17
-------
TABLE 12. SOURCE COAL AUDIT - 0383
Expected
value
0.74%
2.98%
2.05%
6.13%
7.40%
11.42%
12719. Btu/lb
14165. Btu/lb
No. of
analyses
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
125
120
127
120
125
120
127
120
123
118
125
118
125
118
123
118
Laboratories
accurate within i 5%
Sulfur
64%
64%
74%
71%
Moisture
34%
40%
65%
70%
Ash
96%
97%
92%
92%
Gross Calorific
94%
95%
99%
99%
Laboratories
accurate within ± 10%
87%
83%
90%
86%
69%
68%
87%
86%
99%
99%
100%
100%
99%
99%
100%
100%
18
-------
TABLE 13. SOURCE COAL AUDIT - 0983
Expected
value
0.74%
2.95%
1.62%
5.75%
7.31%
11.33%
12470. Btu/lb
14059. Btu/lb
No. of
analyses
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
112
109
113
110
113
110
114
111
112
109
113
110
112
109
111
108
Laboratories
accurate within ± 5%
Sulfur
54%
56%
68%
63%
Moisture
29%
28%
26%
29%
Ash
97%
96%
96%
95%
Gross Calorific
97%
98%
99%
99%
Laboratories
accurate within ± 10%
76%
77%
89%
88%
48%
53%
74%
78%
99%
99%
99%
98%
100%
100%
100%
100%
19
-------
SECTION 7
METHOD 3 AUDIT
This audit checks the participant's ability to analyze a gas sample
using an Orsat analyzer. The audit package consists of a disposable cylin-
der that contains a 4 liter sample of oxygen, carbon dioxide, and carbon
monoxide. The analyst expels the gas into the Orsat analyzer using posi-
tive pressure of the cylinder. The gas sample is analyzed for percentage
of carbon dioxide, oxygen, and carbon monoxide.
In the spring audit (0583), 55% of the 110 laboratories receiving the
audit package returned data. In the fall audit (1183), 53% of the 94
laboratories returned data. Thirty-seven laboratories participated in both
audits and returned data. Table 14 shows the total number of laboratories
requesting participation versus the number that returned data for the Method
3 audits 0583 and 1183.
TABLE 14. METHOD 3 AUDIT PARTICIPANTS
No. requesting samples
Total
110
94
No. returning data
Contractor
Industry
Foreign
Federal
State
Local
0583
66
29
1
2
9
3
1183
52
24
1
2
12
3
0583
30
18
1
2
8
2
1183
25
13
1
1
7
3
61
50
Tables 15 and 16 summarize the Method 3 audits. Each laboratory was
asked to analyze the sample in duplicate. Five and ten percent accuracy
were chosen for the precision reporting criteria for each of the parame-
ters. Each parameter had only one concentration.
In the 0583 audit, only 43% of the reporting laboratories achieved an
accuracy within 5% for the C02. However, since the C02 concentration was
higher in the 1183 audit, 61% of the laboratories achieved an accuracy with-
in 5%. In both audits, 81% of the laboratories achieved an accuracy within
5% of the true value for the ©2 analysis which was three times better than
20
-------
the previous audit (4). For the CO analysis, only 25% of the laboratories
in 0583 and 29% in 1183 achieved an accuracy within 5% and 22% of the
laboratories did not report a value for CO.
TABLE 15. SOURCE METHOD 3 AUDIT - 0583
Expected
value
5.00%
15.10%
1.00%
Expected
value
7.10%
11.20%
0.50%
No. of
analyses
(1) 64
(2) 63
(1) 64
(2) 63
(1) 53
(2) 53
TABLE
No. of
analyses
(1) 52
(2) 51
(1) 52
(2) 51
(1) 40
(2) 40
Laboratories
accurate within ± 5%
O>2
44%
43%
02
84%
84%
CO
32%
21%
16. SOURCE METHOD 3 AUDIT
Laboratories
accurate within ± 5
C02
62%
61%
02
81%
84%
CO
35%
23%
Laboratories
accurate within ± 10%
73%
76%
91%
86%
42%
43%
- 1183
Laboratories
accurate within ± 10%
92%
92%
90%
90%
35%
23%
21
-------
REFERENCES
1. U.S. Environmental Protection Agency. Standards of Performance for
New Stationary Sources - Appendix A. Title 40, Part 60, Code of Fed-
eral Regulations.
2. Fuerst, R.G., E.W. Streib, and M.R. Midgett. A Summary of the EPA
National Source Performance Audit Program - 1980. EPA-600/4-81-077,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, December 1981.
3. Streib, E.W., R.G. Fuerst, and M.R. Midgett. A Summary of the 1981
EPA National Performance Audit Program on Source Measurements. EPA-
600/4-83-026, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, June 1983.
4. Streib, E.W. and M.R. Midgett. A Summary of the 1982 EPA National Per-
formance Audit Program on Source Measurements. EPA-600/4-83-049, U.S.
Environmental Protection Agency, Research Triangle Park, North Caro-
lina 27711, April 1984.
5. Chauvenet, W. Manual of Spherical and Practical Astronomy: Volume
II - Theory and Use of Astronomical Instruments (Method of Least
Squares). J.P. Lippincott and Co., Philadelphia, Pennsylvania,
1863 (?).
6. American Society for Testing and Materials. Annual Book of ASTM Stan-
dards - 1979. Part 26. 01-026079-13, American Society for Testing
and Materials, Philadelphia, Pennsylvania.
22
-------
APPENDIX A
FREQUENCY DISTRIBUTIONS
23
-------
DGM FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCE - 0383
1 Samp.
763
757
754
743
739
737
Mln.
.0
.0
.0
.0
.0
.0
10%
.4
.4
.4
.4
.4
.4
20%
.8
.8
.8
.8
.8
.8
30%
1.2
1.2
1.2
1.2
1.2
1.2
40*
1.6
1.6
1.6
1.5
1.5
1.5
50%
2.2
2.2
2.2
2.2
2.2
2.2
60%
2.9
2.9
2.8
2.7
2.7
2.7
70%
3.7
3.7
3.6
3.6
3.6
3.5
80%
4.4
4.4
4.4
4.2
4.2
4.2
90%
5.9
5.8
5.7
5.6
5.6
5.6
Max.
466.7
60.9
14.7
10.1
9.3
8.2
Mean
5.7
3.0
2.8
2.6
2.6
2.6
Std.
dev.
32.7
4.3
2.3
2.0
2.0
1.9
Skewness
12.39
.00
-.01
-.02
-.04
-.05
Median
2.2
2.2
2.2
2.2
2.2
2.2
10
.£
DGM FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCE - 0983
1 Samp.
614
611
609
597
590
589
Min.
.0
.0
.0
.0
.0
.0
10%
.3
.3
.3
.3
.3
.3
20%
.7
.7
.7
.7
.7
.7
30%
1.1
1.1
1.0
1.0
1.0
1.0
40%
1.5
1.5
1.5
1.5
1.5
1.5
50%
1.9
1.9
1.9
1.9
1.9
1.9
60%
2.4
2.3
2.3
2.3
2.2
2.2
70%
2.9
2.9
2.9
2.8
2.8
2.8
80%
3.5
3.4
3.4
3.3
3.3
3.3
90%
5.1
5.0
5.0
4.8
4.7
4.7
Max.
303.4
69.2
14.4
10.0
8.0
7.2
Mean
4.1
2.6
2.4
2.2
2.2
2.1
Std.
dev.
21.3
3.7
2.3
1.8
1.6
1.6
Skewness
13.57
.02
-.02
-.03
-.05
-.08
Median
1.9
1.9
1.9
1.9
1.9
1.9
-------
S02 FREQUENCY DISTRIBUTION OF PERCENT DIFFERENCE - NO OUTLIERS REMOVED
AUDIT 0283
Sample
No.
1
2
3
4
5
No.
Obs.
113
113
113
112
113
Min.
.06
.00
.01
.03
.00
10%
.35
.24
.30
.19
.27
20%
.82
.50
.36
.52
.55
30*
1.34
.84
.69
.80
.98
40%
1.75
1.39
1.21
1.20
1.40
50%
2.39
1.99
1.56
1.91
1.69
60%
3.21
2.53
2.08
2.19
2.46
70%
4.28
3.49
3.14
2.73
3.38
80%
5.89
4.35
3.86
3.83
4.07
90%
9.97
8.03
6.65
5.79
6.48
Max.
296.27
292.08
296.01
294.96
303.47
Mean
8.76
7.45
6.89
6.62
7.23
Std.
Dev.
31.02
29.94
30.02
29.93
30.85
S02 FREQUENCY DISTRIBUTION OF PERCENT DIFFERENCE - NO OUTLIERS REMOVED
AUDIT 0883
Sample
No.
1
3
4
6
8
No.
Obs.
90
90
90
90
90
Min.
.10
.00
.00
.01
.02
10%
.35
.11
.16
.33
.20
20%
.71
.43
.52
.54
.39
30%
.86
.59
.79
.81
.72
40%
1.03
1.13
1.14
1.27
1.04
50%
1.66
1.49
1.33
1.47
1.28
60%
2.17
1.98
1.60
1.91
1.48
70%
3.43
2.53
2.13
2.91
2.03
80%
4.94
3.70
4.52
3.66
3.51
90%
12.03
8.38
8.25
5.94
8.71
Max.
97.00
96.98
96.97
96.97
96.97
Mean
4.96
4.74
4.91
4.29
3.77
Std.
Dev.
11.86
12.95
13.30
12.03
10.69
-------
NOX FREQUENCY DISTRIBUTION OF PERCENT DIFFERENCE - NO OUTLIERS REMOVED
AUDIT 0483
Sanple
No.
1
2
3
4
5
No.
Obs.
59
62
63
62
63
Min.
.17
.33
.00
.12
.04
10%
.84
.77
.40
.42
.77
20*
2.51
1.37
.59
1.23
1.99
30*
4.02
2.01
1.63
2.01
2.34
40%
6.20
3.45
3.39
3.40
3.08
50%
9.38
4.45
5.69
4.45
4.53
60%
11.39
5.79
7.28
7.84
5.83
70*
14.07
8.54
8.83
11.10
10.88
807.
22.95
13.46
15.66
16.18
15.71
90%
66.16
39.57
31.41
66.48
67.13
Max.
772.70
638.70
887.23
669.47
663.16
Mean
36.91
28.64
32.70
30.91
32.64
Std.
Dev.
110.70
90.85
118.83
95.03
102.22
NOX FREQUENCY DISTRIBUTION OF PERCENT DIFFERENCE - NO OUTLIERS REMOVED
AUDIT 1083
Saaple
No.
2
3
5
7
9
No.
Obs.
53
53
54
54
52
Mln.
.34
.09
.02
.17
.11
10%
.59
.23
.20
.71
.75
20%
1.60
1.56
.93
1.08
1.36
30%
2.96
2.29
1.57
2.12
2.51
40%
3.72
3.39
2.21
3.20
3.19
50*
5.15
4.35
2.85
4.04
5.80
60%
7.09
4.90
4.10
4.74
7.08
70%
11.49
8.01
5.94
5.29
8.71
80%
15.63
11.72
8.40
8.61
11.65
90%
39.78
53.16
21.76
16.77
21.94
Max.
98.14
111.03
155.40
120.89
142.66
Mean
15.32
14.92
13.95
13.69
15.27
Std.
Dev.
24.39
26.90
31.62
28.76
30.32
-------
NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES
STUDY: 0383
Sanple
1000
6000
1000
6000
1000
6000
1000
6000
No.
245
247
245
247
241
243
241
243
Min.
.00
.00
.00
.00
.00
.00
.01
.00
10%
.00
.67
1.46
.82
.14
.09
.04
.08
20%
1.35
1.01
2.44
1.47
.27
.26
.09
.13
30%
1.35
1.68
3.90
2.28
.54
.44
.13
.19
40%
2.70
2.35
5.37
2.94
.81
.61
.17
.28
50% 60%
SULFUR
2.70 4.05 5
3.02 3.36 4
MOISTURE
6.83 8.29 10
3.59 4.40 5
ASH
1.08 1.35 1
.70 .96 1
GROSS CALORIFIC
.24 .31
.37 .50
70%
.41
.36
.24
.22
.62
.23
.48
.72
80%
8.11
6.04
14.63
6.85
2.03
1.49
.73
1.03
90%
10.81
10.74
24.39
12.40
3.11
2.71
1.57
2.46
Max.
24.32
35.23
347.80
38.01
57.97
8.49
7.36
12.66
Mean
5.01
4.37
12.76
5.71
1.88
1.31
.59
1.01
Std.
Dev.
5.30
4.82
32.22
6.86
5.32
1.72
.99
1.84
-------
NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES
STUDY: 0983
Sample
2000
9000
2000
to 9000
oo
2000
9000
2000
9000
No.
223
221
225
223
223
221
221
219
Min.
.00
.00
.17
.00
.00
.00
.01
.01
10%
.68
.00
1.57
1.85
.26
.27
.30
.14
20*
1.36
1.35
3.30
3.70.
.44
.41
.59
.22
30%
1.69
2.70
5.04
5.56
.62
.55
.80
.28
40%
2.71
2.70
6.61
7.41
.88
.68
.96
.36
50% 60%
SULFUR
3.05 4.07
4.05 5.41
MOISTURE
7.30 8.52
9.88 12.96
ASH
1.06 1.32
.82 1.09
GROSS CALORIFIC
1.09 1.24
.43 .52
70%
5.42
8.11
9.57
16.05
1.59
1.23
1.49
.62
80%
6.44
13.51
10.78
21.60
1.94
1.78
1.87
.75
90%
10.51
16.22
12.70
29.63
2.82
2.60
2.45
1.10
Max.
44.41
277.03
87.83
91.98
242.54
484.40
9.59
9.17
Mean
4.90
9.31
8.40
15.12
3.54
5.49
1.39
.63
Std.
Dev.
6.00
26.92
9.07
16.83
22.11
45.38
1.26
.87
-------
NATIONAL ORSAT AUDIT FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES
STUDY: 0583
Std.
Sample No. Min. 101 20% 30% 40Z 50% 60Z 70% 80% 90% Max. Mean Dev.
C02
2000 127 .00 .00 2.00 4.00 4.00 6.00 8.00 8.00 14.00 24.00 98.00 11.15 16.00
<>2
ro
^2000 127 .00 .66 .66 .66 .66 1.99 1.99 2.65 3.31 10.60 35.10 3.98 6.98
CO
2000 106 .00 .00 .00 10.00 10.00 20.00 20.00 40.00 50.00 80.00 80.00 26.13 26.57
-------
NATIONAL ORSAT AUDIT FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES
STUDY: 1183
Std.
Sample No. Min. 10Z 20Z 30% 40% 50Z 60Z 70Z 802 90% Max. Mean Dev.
C02
3000 103 1.41 1.41 1.41 2.82 4.23 4.23 4.23 7.04 7.04 9.86 43.66 5.87 5.88
02
w 3000 103 .00 .00 .00 .89 1.79 1.79 2.63 3.57 4.46 8.93 73.21 4.43 10.01
o
CO
3000 80 .00 .00 .00 20.00 20.00 20.00 20.00 40.00 60.00 60.00 300.00 32.25 43.57
-------
APPENDIX B
INSTRUCTIONS FOR EPA AUDIT MATERIALS
31
-------
INSTRUCTIONS FOR USE OF ENVIRONMENTAL PROTECTION AGENCY
METHOD 5 DRY GAS METER PERFORMANCE TEST DEVICE
NOTE: All procedures referred to are from revised Method 5 published in the
Federal Register, Volume 12, Number 160, Part II, Thursday, August 18,
1977, pp. 41776-41782 and references contained therein. This revised
method should be adhered to in all details in the use of this quality
assurance performance device.
EQUIPMENT: The participant in this study should possess the following
equipment, including the performance test device supplied by EPA.
Quantity Item
1 Method 5/Source Sampling Meter Box
1 Stopwatch, preferably calibrated in decimal minutes
1 Thermometer, ambient range
1 Barometer. If unavailable, call nearest National Weather Service and
request the ABSOLUTE barometric pressure. (Corrected for temperature
and acceleration due to gravity, but not corrected for altitude.)
1 Performance Test Device. A calibrated flow orifice housed in a quick-
connect coupling and identified with an engraved three-digit serial
number.
WARNING: THE DEVICE MUST NOT BE DISASSEMBLED UNDER ANY CIRCUMSTANCES.
Use these devices at room temperature.
PROCEDURE:
Calibration of Vacuum Gauge The vacuum pressure gauge on the
meter box must be calibrated in the range of use (11-22" Hg) against
a standard (Hg Manometer) to ensure accurate results.
Remove the performance test device from its case and insert it into
the gas inlet quick-connect coupling on the source sampling meter box,
Turn the power to the meter box on and start the pump.
Adjust the coarse flow rate control valve and the fine flow rate
control valve to give a reading of 19" Hg (vacuum reading).
CAUTION: The vacuum reading must be accurate and stable for the
test period.
Allow the orifice and source sampling meter box to warm up for 45
minutes with flow controls adjusted as described in Step 3 before
starting quality assurance runs.
32
-------
PROCEDURE: (continued)
Make triplicate quality assurance runs. For each run, record
initial and final dry gas neter volumes, dry gas meter inlet and
outlet temperatures, internal orifice pressure drop (AH), ambient
temperature, and barometric pressure. Run duration should be
lightly greater than 15 minutes. The following procedure is
recommended. Fifteen minutes after a run is started, the partici-
pant watches the dry gas meter needle closely. As the needle reaches
the zero (12 o'clock) position, the pump and stopwatch are stopped
simultaneously. The dry gas meter volume and time are recorded.
This complete run procedure is performed three times to provide the
required triplicate quality assurance runs.
Calculate the corrected dry gas volume for each run using equation 5.1
of the above-referenced Method 5. For each replicate, record the
corrected dry gas volume in dry standard cubic meters, the sampling
time in decimal minutes, the barometric pressure in mm Hg, and the
ambient temperature in degrees Celcius on the enclosed data card. Be
sure to record the performance test device serial number on the data
card in the column headed "Orifice Number."
NOTE 1: If you calculate dry gas volume in English Units, use the
conversion factor of 0.02832 m^ft^ to obtain the volume in
etric units.
NOTE 2: If your stopwatch is not in decimal minutes, be sure to
convert (e.g. 15 minutes 20 seconds is reported as 15.33
minutes).
After recording the requested data on the enclosed data form, return
the data form and the performance test device to:
Quality Assurance Division (MD-77A)
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ATTN: Ellen W. Streib
A postpaid return envelope and label are enclosed for this purpose.
33
-------
INSTRUCTIONS FOR USE OF ENVIRONMENTAL PROTECTION AGENCY STATIONARY
SOURCE QUALITY ASSURANCE SOg REFERENCE SAMPLES
HOTE: All Method 6 procedures referred to are from the amended method pub-
lished in the Federal Register Vol. 1*2. No. 160, Part II, Thursday,
August 16, 1977, pp 1»1782-U1781*. This amended method should be
adhered to in all details in the analysis of these reference stand-
ards.
1. Prepare 3 percent hydrogen peroxide according to Section 3.1.3 of the
method (30 ml is required for each sample and each blank).
2. Prepare each reference sample for analysis as follows: Wrap a paper
towel around the ampoule and with the ampoule in an upright position
break off the top at the prescored mark by exerting pressure sideways.
From the ampoule pipette exactly 5 ml of the reference sample into a
100 ml volumetric flask. Add 30 ml of 3 percent hydrogen peroxide
solution. Dilute exactly to the mark with deionized, distilled water.
Analyze the sample in accordance with the procedure detailed in Section
1».3 of the method, beginning with "Pipette a 20 ml aliquot of this
solution..." (Note: If more than 50 ml of barium perchlorate titrant
is required for any sample analysis, a smaller aliquot should be
selected to allow titration with less than 50 ml titrant.)
3. Calculate the concentration, CsQ2 (concentration of sulfur dioxide,
dry basis, corrected to standard conditions, mg/dscm), using Equation
6-2. A value of 21 x 10"^ dscm should be used for Vm(std), in the
equation. A value of 100 ml should be used for Vsoinin the equation.
1*. Record the reference standard sample numbers and their corresponding
S02 concentrations in mg/dscm on the enclosed data form. Return the
form to:
Quality Assurance Division (MD-77A)
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ATTN: Ellen W. Streib
If other than EPA Method 6 is used for your analyses, please explain in
detail your analytical procedure on the back of the enclosed data form.
-------
INSTRUCTIONS FOR USE OF ENVIRONMENTAL PROTECTION AGENCY STATIONARY
SOURCE QUALITY ASSURANCE NOX REFERENCE SAMPLES
Note: All Method 7 procedures referred to are from the amended method
published in the Federal Register Vol. 42, No. 160, Part 11,
Thursday, August 18, 1977, pp 41784-41786. This amended method
should be adhered to in all details in the analysis of these
reference standards.
1. Prepare absorbing solution according to Section 3.1 of the method.
2. Prepare each reference sample for analysis as follows: Wrap a paper
towel around the ampule and with the ampule in an upright position
break off the top at the prescored mark by exerting pressure sideways.
From the ampule pipette exactly 5 ml of the reference sample into a
100-ml beaker. Add 25 ml absorbing solution to the beaker; adjust
the pH to 9-12 (using pH paper as indicated in Section 4.2 of the
method) by dropwise addition of sodium hydroxide (1N). Quantitatively
transfer the contents of the beaker to a 50-ml volumetric flask and
dilute exactly to the mark with deionized, distilled water. Mix
thoroughly and pipette a 25-ml aliquot of the diluted sample into
a porcelain evaporating dish. Beginning with the evaporation step
in Section 4.3, complete the sample analysis.
3. Calculate total pg NO2 per sample using Equation 7-3. Calculate
the sample concentration, C (concentration of NOX as NO2, dry basis,
corrected to standard conditions, mg/dscm), using Equation 7-4.
A value of 2000 ml should be used for V8C in Equation 7-4.
4. Record the reference sample numbers and their corresponding concentrations,
C, in mg/dscm on the enclosed data form. Return the form to:
Quality Assurance Division (MD-77A)
Environmental Monitoring Systems Laboratory
Environmental Protection Agency
Research Triangle Park, NC 27711
ATTN: Ellen W. Streib
If other than EPA Method 7 is used for your analyses, please explain in detail
your analytical procedure on the back of the enclosed data form.
35
-------
COAL AUDIT PROGRAM INFORMATION
1. There 1s approximately 50 grains of 60 mesh coal per bottle.
2. Analyze the coal samples for moisture and on a dry basis for ash,
sulfur and gross calorific value. Report moisture, ash, and sulfur
In weight percent with gross calorific value reported as BTU/lb.
3. All methods used 1n the analysis of these coal samples should follow
American Society for Testing and Materials (ASTM) recommended procedures
or an accepted automatic analytical device.
f. Suggested procedures are:
Holsture D-3173
Ash D-3174
Sulfur D-3177
Gross Calorific Value ... D-2015
Please note on the data card (columns 17-32) the ASTM method number.
If an ASTM method was not used for analysis note that on the back of
the data card. Be parameter specific.
5. If you cannot analyze the coal sample for all four parameters, analyze
for what you can. Analysis of moisture 1s necessary to calculate on
a dry basis any of the other three parameters. Analysis of sulfur 1s
also necessary for the calculation of gross calorific value.
6. Analyze each sample 1n duplicate (if possible) and record results as
analysis 1 and analysis 2 for each parameter.
7. Most laboratories will use site number 001. Multiple site numbers
are used by laboratories that receive more than one set of samples.
These central laboratories have requested auditing of their satellite
laboratories.
8. After recording the requested data on the enclosed data card, return the
data card to:
Ms. Ellen W. Streib
Quality Assurance Division (MD-77)
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
A postpaid return envelope 1s enclosed for this purpose.
9. If you have any questions concerning this or any source method audit,
please call (919/541-7834).
36
-------
INSTRUCTIONS FOR USING EPA METHOD 3 AUDIT MATERIALS
Equipment Supplied with Audit Kit
(1) Small gas cylinder containing four liters of gas
(2) Nozzle for cylinder (taped on cylinder)
(3) Tygon tubing
Equipment to be Supplied by Participant
(1) Orsat analyzer
(2) Clamp
(3) Extra Tygon or surgical tubing
Procedure
(1) Leak-check apparatus by clamping off tubing.
SEE DIAGRAM.
(2) Fill up burette with gas by using positive pressure. Vent this sample
through the manifold to the atmosphere.
(3) Repeat Step 2.
(4) Fill up burette past fill mark with the gas and carefully vent out
excess to the atmosphere, until the fill mark is reached.
(5) Analyze for C0?, 09 and CO as described in Sections 4.2.5, 4.2.6
and 4.2.7 of EPA Method 3.
(6) Record the results on the data card enclosed with the sample.
(7) Repeat Steps 4 through 6.
CAUTION: If the tubing is punctured excess times, leakage can occur.
Replace if necessary.
Send the data card to the address below. (The cylinder gas can should
not be returned.)
Ms. Ellen Streib
Quality Assurance Division (MD-77A)
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
37
-------
Note: Site number will always be 001 except when other Orsat apparatus
or participants are using the same gas sample. The extra
apparatus or participants should be labeled 002, 003, etc.
APPARATUS SET-UP
38
U.S. GOVERNMENT PRINTING OFFICE' 1985-559-111/10763
------- |