United States
Environmental Protection
Agency
Environmental Monitoring Systems EPA/600/4-86/005
Laboratory February 1986
Research Triangle Park, NC 27711
Research and Development
vvEPA
A Summary of the
1984 EPA National
Performance Audit
Program on Source
Measurements
-------
EPA/600/4-86/005
A SUMMARY OF THE 1984 EPA NATIONAL PERFORMANCE AUDIT PROGRAM
ON SOURCE MEASUREMENTS
E. W. Streib, T. J. Logan and M. R. Midgett
Quality Assurance Division
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
Disclaimer
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
-------
ABSTRACT
In 1984, the Quality Assurance Division conducted 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 ($02) 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 audit, the mean value for all participants differed by
3.1% from the expected (EPA) value. For the Method 6 audit, the average
mean differed by 3.8% from the expected value. The average mean in the
Method 7 audit was 3.2% from the expected value.
In the two coal audits, the parameters measured were sulfur, moisture,
ash, and Btu content. On the average for the sulfur analysis, 82% of the
participants measured within 10% of the expected value; for Btu, 97% of the
participants measured within 10% of the expected value.
In the Method 3 audit, each parameter had only one concentration. The
mean for C02 was 1.5% from the expected value. In this audit, the mean for
02 was 2.1% from the expected value.
This report includes the results of the performance audits done during
the period from January through December 1984.
iv
-------
CONTENTS
Page
Foreword ...............................
Abstract ............................... iv
Figures ............................... vi
Tables ................................ vii
Acknowledgment ............................ viii
1. Introduction ........................ 1
2. Summary ........................... 1
3. Method 3 Audit ....................... 4
4. Method 5 Dry Gas Meter Audit ................ 6
5. Method 6 Audit ....................... 12
6. Method 7 Audit ....................... 13
7. Method 19A Audit ...................... 15
References .............................. 19
Appendices
A. Frequency Distributions ................... 20
B. Instructions for EPA Audit Materials ............ 26
-------
FIGURES
Number Page
1 Cumulative accuracy of participants in 9
Method 5 audit, 0684
2 Laboratories within 2 percent accuracy 10
3 Results of Method 5 audit, 0684 11
vi
-------
TABLES
Number Page
1 Participants' Results from Method 3 Audit 2
(All Data - No Outliers Removed)
2 Participants' Results from Method 5 Audit 2
(All Data - No Outliers Removed)
3 Participants' Results from Methods 6 and 7 Audits 3
(All Data - No Outliers Removed)
4 Participants' Results from Method 19A Coal Audits 4
(All Data - No Outliers Removed)
5 Method 3 Audit Participants 5
6 Source Method 3 Audit • 0784 6
7 Method 5 Audit Participants 7
8 Previous Results of Method 5 Audit 7
9 Method 6 Audit Participants 12
10 Summary of Source S02 Audit 13
11 Method 7 Audit Participants 14
12 Summary of Source NOX Audit 14
13 Coal Audit Participants 16
14 Source Coal Audit - 0384 17
15 Source Coal Audit - 0984 18
vii
-------
ACKNOWLEDGMENT
We express our appreciation to the laboratories that participated in
the National Performance Audit Program for Stationary Sources. Thanks also
to the Standards Laboratory/Performance Evaluation Branch/Environmental
Monitoring Systems Laboratory (EMSL), who did the acceptance testing of the
audit samples, and to the programmers of the Monitoring and Assessment
Division/EMSL for providing the data systems to store and evaluate the data.
viii
-------
Section 1
INTRODUCTION
The Environmental Protection Agency's Environmental Monitoring Systems
Laboratory (EMSL) at Research Triangle Park, NC, established an audit
program in 1977 to evaluate the performance of companies that conduct
compliance testing using EPA reference methods. The audits check the
participants' analytical accuracy in applying the analytical phase 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 performance to other laboratories conducting similar measure-
ments.
Source Test Methods 3, 5, 6, and 7 were each audited once and Method
19A twice in 1984. Each participating laboratory received an audit package
consisting of the audit sample, a data card, instructions, and an envelope
for returning the data to EPA. A label for returning the audit device was
included with the Method 5 audit package. Participants had eight weeks to
use the audit material and return their data to EPA. At the end of this
period all data received were statistically analyzed to determine the
precision (repeatability or reproducibility) and accuracy with respect to
the EPA expected value obtained by the participants. (See Appendix A.)
The Quality Assurance Division of EMSL maintains a repository of audit
samples for the EPA Methods 3, 6, 7, and 19A. These samples are available
to source testing laboratories for such purposes as training new personnel
or conducting quality control checks. Since the expected values for these
samples are included with the analysis instructions, there is no requirement
for the data to be returned to EPA. We recommend that source testing
laboratories use this sample repository to help improve their overall
analytical performance.
This report summarizes the results obtained in the 1984 source audits.
Section 2
SUMMARY
In 1984, 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 laboratories, as well as
local, State, and federal agencies participated.
The results of the 1984 Method 3 audit are summarized in Table 1.
Participants analyzed the gas sample twice for percentage of carbon dioxide,
oxygen, and carbon monoxide. The mean values of carbon dioxide and oxygen
differed by 2* from the expected value. In contrast, the mean values for
carbon monoxide differed as much as 21% from the expected value.
-------
TABLE 1. PARTICIPANTS' RESULTS FROM METHOD 3 AUDIT
(ALL DATA - NO OUTLIERS REMOVED)
Type of
Sample
Small
Cylinder
(Gas)
Audit
Date
0784
Parameter
X CO?
X 02
% CO
No. of
Analyses
64
62
64
62
51
50
Repli-
cate
1
2
1
2
1
2
EPA
(True)
Value
8.00
8.00
9.90
9.90
0.70
0.70
Participant Results
Mean
7.88
7,91
10.11
10.06
0.55
0.57
Median
7.80
7.80
10.00
10.00
0.60
0.60
Std. Dev.
0.65
0.59
1.22
1.21
0.19
0.19
One Method 5 audit was conducted in 1984. The overall results (no
outliers removed) are summarized in Table 2. The mean for all participants
was 3.1% from the expected value. After correcting for outliers, the mean
was 2.2%. The participants' performances did improve from previous audits
in that the standard deviation greatly decreased (2, 3, 4).
TABLE 2. PARTICIPANTS' RESULTS FROM METHOD 5 AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of Audit No. of
Sample Parameter Date Analyses
Mean Median Std.
(X from EPA Value) Dev.
Orifice
Volume
0684
631
3.1
1.9
4.9
Table 3 presents the data (no outliers removed) from the Methods 6 and
7 audits. In the Method 6 audit, the procedure requires the participants to
determine the sulfate content in five aqueous solutions using the titration
procedure. For each concentration, the mean of the participants was 5%
higher tharr the expected value. In contrast, the median differed by IX. In
the Method 6 audit, 50 to 70X of the participants achieved an accuracy
within 2X Of the expected value. Also, the standard deviation is propor-
tional (*30X) to the mean level of the sulfate samples.
-------
TABLE 3. PARTICIPANTS' RESULTS FROM METHODS 6 AND 7 AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of
Sample
Aqueous
Sulfate
Aqueous
Nitrate
Audit
Date
0584
0484
No. of
Analyses
92
91
92
92
92
71
70
70
71
70
EPA True
Value
(mg/DSCM)
282.2
297.4
899.9
968.5
1197.3
91.0
102.4
284.4
466.4
523.3
Participant Results
Mean
296.3
312.7
924.3
999.2
1234.8
96.3
111.1
288.0
464.9
523.2
Median
284.0
299.8
890,3
966.8
1192.5
90.4
100.5
280.0
457.9
517.5
Std. Dev.
83.74
94.59
277.30
301.92
377.38
36.07
49.50
64.95
117.78
104.37
The Method 7 audit procedure requires that the participants determine
the nitrate content in five aqueous solutions. For each concentration, the
mean of the participants was not higher than 8% from the expected value;
but in contrast, the median differed by less than 2%. In this audit, 3li
to 46% of the participants achieved an accuracy within 3% of the expected
value for 4 out of 5 samples. However, on the lowest concentration sample
only 21* achieved this accuracy.
Table 4 summarizes the results of the two coal audits that were con-
ducted in 1984. 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
5% of the expected value on both concentrations. An accuracy of 1% was
achieved on the Btu content.
-------
TABLE 4. PARTICIPANTS' RESULTS FROM METHOD 19A COAL AUDITS
(ALL DATA - NO OUTLIERS REMOVED)
Type of
Sample
Coal
Audit
Date
0384
0984
0384
0984
0384
0984
0384
0984
0384
0984
0384
0984
0384
0984
0384
0984
Parameter
% S
% H20
% Ash
Btu/lb
No. of
Analyses
113
100
113
99
114
102
114
101
112
101
112
100
110
95
110
96
EPA
Value
0.45
1.17
3.28
2.93
2.94
1.86
12.01
10.65
9.45
9.74
13.13
12.87
12115.0
12595.0
13135.0
13710.0
Participant Results
Mean
0.45
1.18
3.04
2.84
2.92
1.90
11.85
10.44
9.46
9.90
13.04
12.75
12020.3
12504.3
13068.9
13850.7
Median
0.46
1.17
3.06
2.83
2.95
1.97
12.16
10.79
9.50
9.67
13.13
12.80
12133.0
12631.0
13087.0
13860.0
Std. Dev.
0.23
0.11
0.29
0.22
0.44
0.33
1.52
1.50
0.47
2.32
1.00
1.07
385.99
409.65
105.50
224.92
Section 3
METHOD 3 AUDIT
This audit checks the participants' ability to analyze a gas sample
using an Orsat analyzer. The audit package consists of a disposable
cylinder that contains a 4-liter sample of oxygen, carbon dioxide, and
carbon monoxide. The analyst expels the gas into the Orsat analyzer using
the positive pressure of the cylinder. The gas sample is quantitatively
analyzed for percentage of carbon dioxide, oxygen, and carbon monoxide.
In the 1984 audit, 50% of the 109 laboratories receiving the audit
package returned data. Table 5 shows the total number of laboratories
requesting participation versus the number that returned data for the Method
3 audit.
-------
TABLE 5. METHOD 3 AUDIT PARTICIPANTS
No. Receiving Samples No. Returning Data
Category 0784 0784
Contractor
Industry
Foreign
Federal
State
Local
57
32
1
2
14
3
21
20
1
1
8
3
Total 109 54
Table 6 summarizes the Method 3 audit results. 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 parameters.
Each parameter had only one concentration.
In the 1984 audit, 82% of the reporting laboratories achieved an
accuracy within 5% for the C02, which was a 34% increase from the last
audit (4). Eighty percent of the laboratories achieved an accuracy within
5% of the expected value for the 03 analysis. For the CO analysis, only
18% of the laboratories achieved an accuracy within 5%, and 19% of the
laboratories did not report a value for CO.
-------
TABLE 6. SOURCE METHOD 3 AUDIT - 0784
Expected No. of Laboratories Laboratories
Value Analyses Accurate within ± 5% Accurate within ± 10%
CO
8.00 (1) 64 80% 92%
(2) 62 85% 94%
02
9.90 (1) 64 78% 94%
(2) 62 81% 95%
CO
0.70 (1) 51 18% 18%
(2) 50 10% 10%
Section 4
METHOD 5 DRY GAS METER AUDIT
In the Method 5 audit procedure, the 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 measure-
ments. The participants convert each of the three volumes to cubic meters
at standard conditions using the formula specified in Eq. 5.1 of Method 5
(Appendix A 40 CFR 60) and record them on the data card. Finally, they
return the orifice and the data card to EPA, where the data undergoes
statistical analysis.
In the 1984 audit, 62% of the 155 laboratories that received the audit
package returned data. This percentage was slightly lower than those
responding in previous audits (2, 3, 4). Table 7 shows the categories of
the participants and compares the number of participants who requested
participation in the Method 5 audit with the number who actually returned
data.
-------
TABLE 7. METHOD 5 AUDIT PARTICIPANTS
Category
Contractor
Industry
Foreign
Federal
State
Local
Total
No. Receiving Samples
0684
78
51
4
2
15
5
155
No. Returning Data
0684
45
31
3
2
12
3
96
Figure 1, a cumulative histogram, shows the accuracy obtained by
participants in the Method 5 audit, expressed as the percent difference
from the expected (EPA) value at various levels of accuracy. The Code of
Federal Regulations (1) requires that the dry gas meter be calibrated within
an accuracy of ±2%. Figure 2 shows that 54* of the reporting laboratories
attained this accuracy. These results are an improvement compared to three
years ago, but may now be leveling off. Table 8 is a summary of previous
audits which, like Figure 2, shows improvement over the past three years.
TABLE 8. PREVIOUS RESULTS OF METHOD 5 AUDIT
Audit
0381
0981
0382
0982
0383
0983
0684
Number of
Analyses
738
723
827
769
763
614
631
Mean
(% from
13.6
4.3
7.6
12.5
5.7
4.1
3.1
Median
EPA Value)
2.7
2.5
2.5
2.5
2.2
1.9
1.9
Std Dev
110.2
9.7
39.5
81.4
32.7
21.3
4.9
-------
The histogram in Figure 3 compares the individual results from all
participants of the 1984 audit with the mean and median values. The
majority of the laboratories reported values lower than the EPA value. The
standard deviation of the triplicate analysis (repeatability) by each
laboratory indicated that 75% of the standard deviations for each set were
within 0.3%. Four percent of the 1984 data were identified as outliers
using Chauvenet's Criterion (5). Before the outliers were removed, the
mean value differed by 3.1% from the expected value. After deletion of
outliers, this value was reduced to 2.2%.
8
-------
WikOlO>vlOO(OO
ooooooooooo
m M
3)
m
Z
O
11 fa)
z
2
m
w *
c
31
m
m tn
to
•n
31
O
2
m
m
n
»
09
i—i—i—r
T
T
T
T
O)
w
-------
55
50
*«
»
u
i
(A
HI
£ 45
O
oc
O
m
40
35
DEGREE OF ACCURACY = 2 percent
> *
0381/738 0981/723
0382/827 0982/769 0383/763 0983/614
AUDIT/NUMBER OF SAMPLES
0684/631
Figure 2. Laboratories within 2 percent accuracy.
10
-------
LABORATORIES, percent
^ .4
01 o 01
A
(O
C
(D
30
0)
(A
C_
r+
V)
I
I
01
c
Q.
*•+"
O
O)
00
00
o>
m »
3D
m
Z
% b
3D
O
S o
m
•u
— M
m ,jj,
oo
o
V
N)
1
1
S.
00
m
3D'
S-
§•
m m'
I "-
S O>
,\
m
>
2
ii
•
O
NJ
-------
Section 5
METHOD 6 AUDIT
This audit checks the participants' ability to quantitatively analyze
Method 6 samples for sulfate content. The audit set consists of five
aqueous dilutions of 10 N sulfuric acid in 25 ml sealed glass ampoules.
The analyst withdraws 5 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 (BaCCIO^) to a pink endpoint. In calculating the results, the
participants assume an original sample volume of 100 ml, and a sample volume
of 0.021 dry standard cubic meter of stack gas.
Table 9 shows the categories of the participants and compares the
total number of participants requesting participation with the number
returning data. In the 1984 audit, 59% of the 157 laboratories that received
the audit package returned data.
TABLE 9. METHOD 6 AUDIT PARTICIPANTS
No. Receiving Samples No. Returning Data
Category 0684 0684
Contractor
Industry
Foreign
Federal
State
Local
79
49
5
1
15
8
40
31
3
1
11
6
Total 157 92
Table 10 shows the percentage of laboratories that achieved 2% and 5%
accuracy for each of the five different concentrations in the Method 6
audit. Also, at least 50% of the reporting laboratories achieved an accuracy
within 2% for the three lower concentrations, but showed an improvement of
65% for the higher concentrations. Seventy-five percent of the laboratories
were able to achieve an accuracy level of within 5% on all five samples.
12
-------
TABLE 10. SUMMARY OF SOURCE S02 AUDIT
Concentration
282.2 mg/DSCM
297.4 mg/DSCM
899.9 mg/DSCM
968.5 mg/DSCM
1197.3 mg/DSCM
n
0584
± 2%
50%
48%
55%
63%
70%
± 5%
79%
76%
88%
88%
91%
92
Section 6
METHOD 7 AUDIT
This audit checks the participants1 ability to quantitatively analyze
Method 7 samples for nitrate content. The NOX audit set consists of five
aqueous 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 of solution
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 analyzed as described in Section 4.3 of
Method 7 (1). After this treatment is completed, the absorbance is measured
at 410 nm with a calibrated spectrophotometer. In calculating the concen-
trations present, the participant assumes that 2000 ml of stack gas was
sampled.
Table 11 shows the total number of laboratories requesting participa-
ting in and the number that returned data for the 1984 Method 7 audit. Sixty
percent of the 118 laboratories receiving the audit package returned data.
13
-------
TABLE 11. METHOD 7 AUDIT PARTICIPANTS
Category
Contractor
Industry
Foreign
Federal
State
Local
Total
No. Receiving Samples
0484
66
37
3
1
7
4
118
No. Returning Data
0484
36
23
2
1
5
4
71
The percentage of laboratories that achieved 5 and 10% accuracy for
each of the five concentrations are shown in Table 12. In addition, 40% of
the reporting laboratories achieved an accuracy within 5% on the lowest
concentrations and 55% achieved within 5% on the highest three concentra-
tions. Seventy percent of the 71 laboratories were able to achieve an
accuracy within 10% on all five samples.
TABLE 12. SUMMARY OF SOURCE NOX AUDIT
Concentration
91.0 mg/DSCM
102.4 mg/DSCM
284.4 mg/DSCM
466.4 mg/DSCM
523.3 mg/DSCM
n
0484
± 5%
38%
49%
53%
54%
63%
± 10%
68%
66%
74%
77%
79%
71
14
-------
Section 7
METHOD 19A COAL AUDIT
Standards of performance for newer electric utility steam generators
(Subpart Da of 40 CFR 60) allow coal sampling and analysis to serve as an
acceptable method for determining scrubber inlet flue gas sulfur concen-
tration. The coal audit checks the participants' ability to analyze coal
samples for sulfur, ash, moisture, and Btu content.
Coal audit samples consisted of two bottles, each containing 50
grams of 60-mesh coal. The following American Society for Testing and
Materials (ASTM) procedures (6) were recommended, but not required, for
participants' use in analyzing the coal samples:
' ASTM D-3177 (Standard Test Method for Total Sulfur in the Analysis
of Coal and Coke);
0 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
0 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 (X),
ash (%), and gross calorific value (Btu/16) on a dry basis.
In both audits, 83% of the laboratories that received the audit package
returned data. Seventy-two of the same laboratories participated in both
audits and returned data. Table 13 shows the total number of laboratories
requesting participation versus the number that returned data for coal
audits 0384 and 0984.
15
-------
TABLE 13. COAL AUDIT PARTICIPANTS
No. Receiving Samples No. Returning Data
Category 0384 0984 0384 0984
Contractors
Industry
Federal
State
Local
59
40
0
12
5
41
47
1
13
4
42
38
0
11
5
35
38
0
11
4
Total 116 106 96 88
Tables 14 and 15 summarize the coal audit results. The number of
analyses 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 the 0384 audit, 60% of the laboratories were able to analyze the
sulfur content of the low level sample within 10% of the expected value;
however, 86% were able to analyze the high level sample within 10% of the
expected value. In the 0984 audit, 90% of the laboratories were able to
analyze both sulfur samples within 10% of the expected value. Only 50%
achieved 10% on the low moisture concentration in the 0984 audit; whereas,
75% in the 0384 audit achieved 10% on the low moisture concentration. For
the ash analysis and Btu content, 92% to 100% of the reporting laboratories
were able to achieve an accuracy within 10% for both sample concentrations.
The participants' accuracy improved with higher concentrations on all
parameters except ash in the 0984 audit, where there was a slight decrease.
For those laboratories that reported duplicate analyses, the intra-labora-
tory precision (repeatability) was not dependent on concentration level.
16
-------
TABLE 14. SOURCE COAL AUDIT - 0384
Expected
Value
0.45
0.45
3.28
3.28
2.94
2.94
12.01
12.01
9.45
9.45
13.13
13.13
12115.
12115.
13135.
13135.
No. of
Analyses
(D*
(2)*
(D
(2)
(D
(2)
(D
(2)
(D
(2)
(D
(2)
(D
(2)
(D
(2)
113
100
113
99
114
100
114
101
112
100
112
98
110
99
110
97
Laboratories
Accurate within ± 5%
SULFUR
42*
37%
23%
28%
MOISTURE
48%
51%
79%
81%
ASH
91%
93%
96%
96%
GROSS CALORIFIC
87%
94%
100%
100%
Laboratories
Accurate within ± 10%
60%
60%
86%
86%
75%
78%
92%
93%
94%
94%
98%
99%
89%
95%
100%
100%
* Indicates first and second analyses
17
-------
TABLE 15. SOURCE COAL AUDIT - 0984
Expected
Value
1.17
1.17
2.93
2.93
1.86
1.86
10.65
10.65
9.74
9.74
12.87
12.87
12595.
12595.
13710.
13710.
No. of
Analyses
(D*
(2)*
(D
(2)
(D
(2)
(1)
(2)
(D
(2)
(D
(2)
(D
(2)
(D
(2)
100
95
99
95
102
96
101
96
101
95
100
95
95
91
96
91
Laboratories
Accurate within ± 5%
SULFUR
77%
77%
64%
63%
MOISTURE
25%
20%
79%
84%
ASH
95%
92%
87%
86%
GROSS CALORIFIC
89%
90%
99%
99%
Laboratories
Accurate within ± 10%
95%
93%
90%
89%
48%
51%
92%
93%
97%
97%
92%
92%
96%
96%
99%
99%
* Indicates first and second analyses
18
-------
REFERENCES
1. U.S. Environmental Protection Agency. Standards of Performance for New
Stationary Sources - Appendix A. Title 40, Part 60, Code of Federal
Regulations.
2. Streib, E. W., R. 6. 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.
3. Streib, E. W. and M. R. Midgett. A Summary of the 1982 EPA National
Performance Audit Program on Source Measurements. EPA-600/4-83-049,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, April 1984.
4. Streib, E. W., T. 0. Logan and M. R. Midgett. A Summary of the 1983
EPA National Performance Audit Program on Source Measurements. EPA-
600/4-85-004, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711, January 1985.
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 Stand-
ards - 1984. Volume 5.05. 01-050584-13, American Society for Testing
and Materials, Philadelphia, Pennsylvania.
19
-------
APPENDIX A
FREQUENCY DISTRIBUTIONS
21
-------
ro
ro
OGM FREQUENCY DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCE - 0684
f SAHP MIN 10% 20* 30% 40% 50% 60% 70% 80% 90% MAX MEAN STDEV SKEWNESS MEDIAN
631
625
611
603
599
598
596
595
593
590
587
585
.0
,0
.0
.0
,0
.0
.0
.0
.0
.0
.0
.0
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.7
,7
.7
.6
,6
.6
.6
.6
.6
.6
.6
.6
1,0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
.9
1.4
1.4
1.3
1,3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.3
1.9
1.8
1.8
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
1.7
2.5
2.4
2.4
2.3
2.3
2,3
2,2
2.2
2.2
2.2
2.2
2.2
3.1
3.1
3.0
2.9
2.9
2.9
2.9
2. a
2.8
2,8
2.8
2.8
4.4
4.2
4.0
3,8
3,8
3.7
3.7
3.7
3.7
3.6
3.6
3.5
6.1
5.9
5.5
5.2
5.2
5.2
5.2
5.1
5.1
5.1
5.1
5.0
58.8
17,1
12.1
9,6
9.1
8.9
8.9
8,7
8.6
8.5
8.2
8.0
3.1
2.8
2.5
2.4
2,3
2.3
2.3
2.3
2.3
2.2
2.2
2.2
4.9
2.9
2.3
2.1
2.0
2.0
2.0
1.9
1,9
1.9
1.8
1.8
6,97
-.12
-.27
-.38
-,48
-.56
-.64
-.71
-.77
-.83
-.88
-.93
1.9
1.9
1,8
1.7
1.7
1.7
1.7
1.7
1,7
1.7
1.7
1.7
-------
ro
CO
S02 FREQUENCY DISTRIBUTION OF PERCENT DIFFERENCE - NO OUTLIERS REMOVED
AUDIT 0584
Sample
No.
2
3
5
7
9
NO.
91
92
92
92
92
MIN
.00
.00
.00
.00
.00
NOX
105!
.27
.43
.36
.26
.20
20%
.74
.67
.54
.61
.45
30%
1.14
1.13
.87
1.09
.61
40%
1.51
1.45
1.09
1.44
.90
FREQUENCY DISTRIBUTION OF
50%
2.15
1.88
1.60
1.77
1.36
PERCENT
60% 70%
2.56 3.56
2.69 3.40
1.87 2.55
2.22 2.63
1.65 1.99
DIFFERENCE -
80%
6.52
4.32
3.25
3.49
3.12
90%
10.26
9.14
5.16
5.01
4.15
MAX
300.13
275.62
296.49
293.38
300.90
MEAN
7.10
7.10
5.69
5.57
5.21
STDEV
31.42
29.24
30.81
30.42
31.24
NO OUTLIERS REMOVED
AUDIT 0484
Sample
No.
1
4
6
8
9
NO.
70
70
71
71
70
MIN
,11
.00
.00
.11
.29 1
10%
.67
.63
.99
.73
.17
20%
1.44
.90
2.75
1.37
2.25
30%
1.83
1.78
4.40
2.79
2.83
40%
2.43
2.35
5.27
3.52
3.81
50%
4.36
3.50
6.92
4.40
5.08
60% 70%
6.12 9.11
4.36 5,33
7.69 10.99
5.66 7.20
7.03 10.64
80%
11.46
10.09
15.27
10.38
15.82
90%
18.74
16.17
31.87
16.66
38.67
MAX
127.74
87.46
278.02
117.13
347.46
MEAN
10.41
9.18
16.83
11.44
18.09
STDEV
20.33
17.67
36,30
22.48
45.59
-------
ro
NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION
OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES - STUDY 0384
SULFUR
Sample
No.
4000
7000
NO.
212
213
MIN
.00
.00
10%
3.35
2.22
20%
4.57
2.22
30%
5.18
4.44
40%
6.10
6.67
50%
6.71
8.89
60%
7.32
8.89
70%
8.23
13.33
80%
9.15
15.56
90%
10.37
24.44
MAX
88.11
535.56
MEAN
7.80
16.03
STDEV
8.47
51.99
MOISTURE
Sample
No.
4000
7000
Sample
No.
4000
7000
Sample
No.
4000
7000
NO.
214
215
NO.
210
212
NO.
207
209
MIN
.00
.00
MIN
.00
.00
MIN
.01
.00
10%
1.36
.58
10*
.15
.21
10%
.05
.06
20%
2.38
1.00
20%
.23
.42
20%
.11
.17
30%
3.40
1.33
30%
.38
.74
30%
.20
.26
40%
4.42
1.75
40%
.53
.95
40%
.27
.34
50%
5.44
2.25
ASH
50%
.61
1.16
GROSS
50%
.40
.40
60%
6.46
2.83
60%
.76
1.59
CAL
60%
.55
.54
70%
8.16
3.50
70%
.99
1.80
70%
.73
.67
80%
10.88
5.00
80%
1.37
2.43
80%
.97
.93
90%
14.97
7.16
90%
2.36
3.28
90%
1.36
2.48
MAX
104.76
97.59
MAX
76.69
32.80
MAX
4.59
13.02
MEAN
7.83
4.38
MEAN
1.54
2.40
MEAN
.65
1.31
STDEV
9.97
10.30
STDEV
5.58
4.39
STDEV
.75
2.84
-------
ro
en
NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION
OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES - STUDY 0984
SULFUR
Sample
No.
5000
8000
Sample
No.
5000
8000
Sample
No.
5000
8000
Sample
No.
5000
8000
NO.
195
194
NO.
198
197
NO.
196
195
NO.
187
186
MIN
.00
.00
MIN
.00
.00
MIN
.00
.00
MIN
.06
.02
10%
.85
.34
10%
2.15
.66
10%
.21
.08
10*
.50
.21
20%
.85
1.02
20%
4.30
1.22
20%
.31
.23
20%
.74
.37
30%
1.71
1.71
30%
6.99
1.88
30%
.51
.47
30%
.85
.52
40%
1.71
2.73
40%
8.06
2.25
40%
.72
.62
40%
.97
.67
50% 60%
2.56 3.42
3.75 4.78
MOISTURE
50% 60%
10.22 11.29
2.72 3.38
ASH
50% 60%
.92 1.13
.85 1.01
GROSS CAL
50% 60%
1.09 1.20
.78 .91
70%
4.27
5.80
70%
13.44
3.94
70%
1.64
1.40
70%
1.26
1.05
80%
5.13
7.17
80%
16.13
4.79
80%
2.05
2.18
80%
1.39
1.71
90%
8.55
9.90
90%
22.04
6.20
90%
3.90
7.23
90%
1.54
4.92
MAX
101.71
44.71
MAX
89.25
90.61
MAX
250.10
113.68
MAX
15.28
12.18
MEAN
4.36
5.06
MEAN
12.78
5.35
MEAN
4.12
3.18
MEAN
1.22
1.84
STDEV
9.62
5.79
STDEV
12.84
12.94
STDEV
24.51
10.22
STDEV
1.53
2.82
-------
NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION
OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES - STUDY 0784
CO
Sample NO. MIN 10X 20* 30* 40* 50* 60* 70* 80* 90* MAX MEAN STDEV
No.
4000 126 .00 .00 1.25 2.50 2.50 2.50 3.75 5.00 5.00 7.50 45.00 4.41 6.53
Sample NO. MIN 10X 20* 30* 40* 50* 60* 70* SOX 90* MAX MEAN STDEV
No.
ro ' '—'
°" 4000 126 .00 1.01 1.01 1.01 1.01 2.02 3.03 3.03 4.04 5.05 75.76 4.74 11.38
CO
Sample NO. MIN 10* 20* 30* 40* 50* 60* 70* 80* 90* MAX MEAN STDEV
No.
4000 101 .00 .00 14.29 14.29 14.29 14.29 28.57 28.57 42.86 57.14 85.71 26.17 21.19
-------
APPENDIX B
INSTRUCTIONS FOR EPA AUDIT MATERIALS
27
-------
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 CO,,, 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
28
-------
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
29
-------
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.
30
-------
PROCEDURE: (continued)
Make triplicate quality assurance runs. Ftor each run, record
initial and final dry gas meter volumes, dry gas meter inlet and
outlet temperatures, internal orifice pressure drop (AH), ambient
temperature, and barometric pressure. Run duration should be
slightly 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
metric 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.
31
-------
INSTRUCTIONS FOR USE OF ENVIRONMENTAL PROTECTION AGENCY STATIONARY
SOURCE QUALITY ASSURANCE S02 REFERENCE SAMPLES
NOTE: All Method 6 procedures referred to are from the amended method pub-
lished in the Federal Register Vol. 1*2, No. l6o, Part II, Thursday,
August 18, 1977, PP 1*1782-1*1784. 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
U.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, 0302 (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 V8Oinin 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.
32
-------
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 yg NO2 per sample using Equation 7-3. Calculate
the sample concentration, C (concentration of NOX as N02, dry basis,
corrected to standard conditions, mg/dscm), using Equation 7-4.
A value of 2000 ml should be used for Vsc 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.
33
-------
COAL AUDIT PROGRAM INFORMATION
1. There 1s approximately 50 grams 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
1n 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.
4. Suggested procedures are:
Moisture 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 Is
also necessary for the calculation of gross calorific value.
6. Analyze each sample In 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 nave any questions concerning this or any source method audit,
please call (919/541-7834).
34
frUS GOVERNMENT PRINTING OFFICE 1966 - 646-116/20766
------- |