United States
           Environmental Protection
           Agency
          Environmental Monitoring Systems
          Laboratory
          Research Triangle Park NC 27711
EPA/600/4-87-001
December 1986
           Research and Development
vvEPA
A Summary of the
1985 EPA National
Performance Audit
Program on Source
Measurements

-------
                                                  EPA/600/4-87/001
                                                  December 1986
A SUMMARY OF THE 1985 EPA NATIONAL PERFORMANCE AUDIT PROGRAM
                   ON SOURCE MEASUREMENTS
        E. W. Streib, T. J. Logan, and M.  R.  Midgett
                 Quality Assurance Division
        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

-------
                       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.

-------
                                    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 Method 3 (carbon dioxide and oxygen), EPA Method  5
(dry gas meter only), Method 6 (sulfur dioxide), Method 7 (nitrogen oxides),
and Method 19 (coal).  Statistical  analysis was used to characterize  the
data.
                                             John  C.  Puzak,  Deputy Director
                                Environmental  Monitoring Systems  Laboratory
                                     Research  Triangle Park, North Carolina
                                    iii

-------
                                  ABSTRACT


     In 1985, the Quality Assurance Division conducted the National  Audits
for Stationary Source Test Methods.  The audit  materials  consisted of:   a
disposable gas cylinder for Method 3 (Orsat analyzer), a  calibrated orifice
for Method 5 (dry gas meter only), five simulated liquid  samples  each for
Method 6 (S02J and Method 7 (NOX), and two coal samples for Method 19A.
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 3 audit, each gas component  had only one concentration.
The mean C02 value for all participants differed by 4.0 percent from the
expected (EPA) value, and the mean for 02 was 0.4 percent from the expected
value.

     In the Method 5 audit, the mean value for  all participants differed by
10.6 percent from the expected value.  For the  Method 6 audit, the average
mean differed by 10 percent from the expected value.  The average mean  in the
Method 7 audit was 4 percent 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, 87 percent of
the participants measured within 10 percent of  the expected value; for  Btu,
98 percent of the participants measured within  10 percent of the  expected value.

     This report includes the results of the performance  audits done during
the period from January to December 1985.
                                     iv

-------
                                  CONTENTS

                                                                        Page
Foreword	   Ill
Abstract 	    1 v
Figures	    vl
Tabl es 	   vll
Acknowledgment	  vl 11
     1.  Introduction 	     1
     2.  Summary 	     2
     3.  Method 3 Audit 	     6
     4.  Method 5 Dry Gas Meter Audit  	     8
     5.  Method 6 Audit 	    13
     6.  Method 7 Audit 	    16
     7.  Method 19A Audit 	    19
References	    23
Appendixes
     A.  Frequency Distributions 	    24
     B.  Instructions for EPA Audit Materials  	    30

-------
                           FIGURES
Number                                                          Page
  1           Cumulative accuracy for participants in
              Method 5 audit, 0685 	  10
  2           Previous results of Method 5 audit	  11
  3           Results of Method 5 audit, 0685 	  12
  4           S02 expected values compared to participants'
              values 	  15
  5           NOX expected values compared to participants'
              values 	  18
                              VI

-------
                                   TABLES
Number                                                                 Page

  1              Participants'  Results from Method 3 Audit
                 (All  Data—No Outliers Removed)  	    2

  2              Method 5 Audit 0685 - Summary Statistics   	    3

  3              Participants'  Results from Methods 6 and  7 Audits
                 (All  Data—No Outliers Removed)  	    3

  4              Participants'  Results from Method 6 and 7  Audits
                 (Outliers Removed)  	    4

  5              Participants'  Results from Method 19A Coal Audits
                 (All  Data—No Outliers Removed)  	    5

  6              Method 3 Audit Participants	    6

  7              Source Method 3 Audit—0785 	    7

  8              Method 5 Audit Participants 	    8

  9              Previous Results of Method 5 Audit 	    9

 10              Method 6 Audit Participants 	   13

 11              Summary of Source SOg Audit—0585 	   14

 12              Method 7 Audit Participants 	   16

 13              Summary of Source NOX Audit—0485 	   17

 14              Coal  Audit Participants 	   19

 15              Source Coal Audit—0385 	   21

 16              Source Coal Audit—0985 	   22

 17              National Orsat Audit Frequency Distribution of
                 Absolute Percent Differences of Expected  and Reported
                 Values—Study 0785  	   25

 18              DGM Frequency Distribution of Absolute Percent
                 Di fference—0685 	   26
                                    VI1

-------
                                   TABLES
Number                                                                Page

  19            S0£ Frequency Distribution of Percent Difference—No
                Outliers  Removed—Audit 0585 	   27

  20            NOX Frequency Distribution of Percent Difference—No
                Outliers  Removed—Audit 0485 	   27

  21            National  Coal Audit  Frequency Distribution of Absolute
                Percent Differences  of Expected and Reported Values-
                Study 0385  	   28

  22            National  Coal Audit  Frequency Distribution of Absolute
                Percent Differences  of Expected and Reported Values-
                Study 0985  	  29

-------
                               ACKNOWLEDGMENT
     We express our appreciation to the laboratories that participated in
the National Performance Audit Program for Stationary Sources.   Thanks also
to the staff of 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.

-------
                                 SECTION 1


                                INTRODUCTION


     The Environmental Protection Agency's (EPA's) Environmental  Monitoring
Systems Laboratory (EMSL) at Research Triangle Park, North Carolina, estab-
lished 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 a 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
measurements.

     Source Test Methods 3, 5, 6, and 7 were each audited once and Method  19A
twice in 1985.   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 8 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 accuracy
with respect to the EPA expected value and those obtained by the participants
(see Appendix A).

     The Quality Assurance Division of EMSL also maintains a limited
repository of samples for the EPA Methods 3, 6, 7, and 19A that are available
to source testing laboratories for such purposes as training new personnel
or conducting quality control checks.  Because 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 1985 source audits.

-------
                                 SECTION 2
                                  SUMMARY

     In 1985, EPA's EMSL at Research Triangle  Park,  North  Carolina,  conducted
National Quality Assurance Audits for Stationary Source Test  Methods 3
(Orsat analyzer), 5 (dry gas meter only), 6 ($02),  7 (NOX), and 19A  (coal).
Industrial laboratories, contractors, foreign  laboratories, as well  as
local, State, and Federal  agencies participated.

     The results of the 1985 audit of Method 3 are  summarized in Table 1.
Participants analyzed the gas sample twice for percentages of carbon dioxide
(C02), oxygen (03), and carbon monoxide (CO).   The  mean values of C02 and
62 differed by 4 percent and 0.4 percent from  the expected values, respectively.
In contrast, the mean values for CO differed as much as 63 percent from the
expected values.
            TABLE 1.  PARTICIPANTS1  RESULTS FROM METHOD 3 AUDIT
                      (ALL DATA—NO OUTLIERS REMOVED)

Type of Audit
sample date
Small 0785
cylinder
(gas)






Parameter
% C02


% 02

% CO


No. of
analyses
46
45

47
46
37
36

Repli-
cate
1
2

1
2
1
2
EPA
true
value
7.00
7.00

12.00
12.00
0.60
0.60



Participant results
Mean
6.76
6.68

11.98
12.07
0.96
1.00
Median
6.60
6.60

12.00
12.00
0.50
0.50
Std.dev.
0.97
0.88

1.38
1.33
2.92
2.97
     One audit of Method 5 was conducted in 1985.  The overall results (no
outliers removed) are summarized in Table 2.  The mean for all participants
was 10.6 percent from the expected value and the standard deviation was
102.1 percent.  After the removal of 7 percent of the data statistical outliers,
the mean was 2.3 percent from the expected values and the standard deviation
was 1.7 percent.  The participants' performance based on the standard deviation
of all data was poor; however, when the 7 percent of the outliers was removed,
the performance was consistent with previous years.

-------
             TABLE 2.  METHOD 5 AUDIT 0685 - SUMMARY STATISTICS
                                         Mean
Median
Std. dev
All data
Outliers removed
633
588
10.6
2.3
2.2
2.0
102.1
1.7
     Table 3 represents the data (no outliers removed)  from the Methods 6
and 7 audits.  In the Method 6 audit, the procedure requires the partici-
pants to determine the sulfate content in five aqueous  solutions using the
titration procedure.  For four out of five concentrations, the mean of the
participants was 10 percent higher than the expected value.  In contrast, the
median differed by 2 percent.  In the Method 6 audit, 60 to 85 percent of
the participants achieved an accuracy within 5 percent  of the expected value.

     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 less than 4 percent from the expected value,
and the median differed by less than 2 percent.  In this audit, 70 percent
of the participants achieved an accuracy within 10 percent of the expected
value for all five concentrations.
        TABLE 3.  PARTICIPANTS'  RESULTS FROM METHODS 6 AND 7  AUDITS
                      (ALL DATA—NO OUTLIERS REMOVED)

Type of Audit
sample date
Aqueous 0585
sulfate



Aqueous 0485
nitrate




No. of
analyses
89
91
90
91
89
72
73
73
72
73
EPA true
value
(mg/DSCM)
198.3
221.2
701.6
1250.7
1326.9
99.5
119.5
298.6
338.4
537.5
Participant results
Mean

243.8
245.5
723.7
1332.6
1426.1
97.9
118.3
293.8
344.8
518.4
Median

201.5
225.0
698.3
1240.1
1323.1
101.1
119.0
298.0
339.4
534.0
Std.
dev.
263.8
104.5
247.9
547.5
595.3
24.4
23.0
49.5
73.8
112.9

-------
     Table 4 represents the data from Methods 6 and 7 after the removal  of
outliers.  Two percent of the Method 6 data and 3 percent of the Method 7
data were classified as outliers.  It is important to note that even with
the statistical outliers removed, the performance of this year's Method 6
data (based on the standard deviation) is not as good as in previous years.
         TABLE 4.  PARTICIPANTS'  RESULTS FROM METHOD 6 AND 7 AUDITS
                             (OUTLIERS REMOVED)
Type of
Sample
Audit
Date
No. of
Analyses
EPA true value
(mg/DSCM)
Participant
Mean
Median
results
Std. dev
Aqueous
 sulfate
0585
88
88
89
89
87
 198
 221
 701
1250
1326
,3
.2
,6
.7
.9
217.6
227.6
                                               699
                                              1252
                                              1338
 201.3
 224.6
 697.8
1240.0
1322.2
 91.9
 25.2
 95.9
103.6
117.0
Aqueous 0485
nitrate



69
71
71
70
71
99.5
119.5
298.6
338.4
537.5
101.5
120.8
299.4
342.9
530.9
101.9
119.1
298.5
339.4
534.0
17.5
17.7
36.3
49.5
84.8
     Table 5 summarizes the results of the two coal  audits that were con-
ducted in 1985.  Participants analyzed each coal  sample in duplicate for
percentages of sulfur, moisture, and ash, and for gross calorific value
(Btu/lb).  The means of the ash, moisture, and sulfur content were within
6 percent of the expected value except for a 20 percent difference on the
lowest sulfur concentration.  An accuracy of 1 percent was achieved on the
Btu content.  The Btu content as measured by the participants is lower than
the EPA value because coal oxidation is not prevented in the audit sample.
The Btu content is a good example of the accuracy to which a measurement can
be made when that measurement receives good laboratory quality control.

-------
TABLE 5.  PARTICIPANTS' RESULTS FROM METHOD 19A COAL AUDITS
              (ALL DATA—NO OUTLIERS REMOVED)
Type of Audit
sample date
Coal 0385
0985
0385
0985
0385
0985
0385
0985
0385
0985
0385
0985
0385
0985
0385
0985
No. of
Parameter analyses
% S 111
108
110
106
% H20 111
106
112
108
% Ash 111
107
110
105
Btu/lb 108 11
102 12
109 13
104 12
EPA
value
1.62
0.72
4.97
2.35
3.86
1.61
3.93
17.45
11.36
4.33
20.32
17.92
,414.0
,300.0
,054.0
,374.0
Participant results
Mean
1.60
0.86
4.75
2.29
3.60
1.75
3.97
16.73
11.37
4.25
19.92
17.59
11,379.7
12,252.9
13,011.1
12,179.3
Median
1.59
0.70
4.76
2.29
3.70
1.53
3.95
17.41
11.33
4.37
20.07
17.83
11,383.5
12,273.0
13,025.0
12,319.0
Std.dev.
0.14
1.56
0.35
0.23
0.52
1.86
1.65
2.61
0.45
0.44
1.52
2.34
256.61
183.77
261.53
573.78

-------
                                 SECTION 3


                               METHOD 3 AUDIT


     The Method 3 audit checks participants'  ability to analyze a gas sample
using an Orsat analyzer.  The audit package consists of a disposable cylinder
that contains a 4-liter (L) sample of COg, Og, and CO.   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 C02, 02, and CO.

     In the 1985 audit, 44 percent of the 103 laboratories receiving the audit
package returned data.  Table 6 shows the total  number  of laboratories
requesting participation and the number that returned data for the Method 3
audit.
	TABLE 6.  METHOD 3 AUDIT PARTICIPANTS	

                             No. receiving                 No.  returning
Category                       samples                          data
Contractors
Industry
Foreign
Federal
State
Local
57
28
1
2
12
3
16
15
1
2
8
3
Total                           103                             45
     Table 7 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 reporting criteria for each of the parameters.  Each
parameter had only one concentration.
     In the 1985 audit, 35 percent of the reporting laboratories achieved
an accuracy within 5 percent for the C02, which was a decrease from the
last audit.'4) Ninety percent of the laboratories achieved an accuracy within

-------
5 percent of the expected value for the 02 analysis.   For the CO analysis,
only 33 percent of the laboratories achieved an accuracy within 5 percent,
and 21 percent of the laboratories did not report a value for CO.
                   TABLE 7.   SOURCE METHOD 3 AUDIT--0785
Expected
value
7.00
12.00
0.60
No. of
analyses*
(1) 46
(2) 45
(1) 47
(2) 46
(1) 37
(2) 36
Laboratories
accurate within 5%
(%)
35
30
89
91
32
36
Laboratories
accurate within 10%
(%)
85
89
92
94
32
36
 * Numbers  in parentheses  indicate first  and  second  analyses.

-------
                                 SECTION 4
                        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-min volume measurements.
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.   Then they return
the orifice and the data card to EPA, where the data undergoes statistical
analysis.

     In the 1985 audit, 59 percent of the 160 laboratories that received  the
audit package returned data.  Table 8 shows the categories of participants,
the number of participants who requested participation  in the Method 5 audit,
and the number who actually returned data.
	TABLE 8.  METHOD 5 AUDIT PARTICIPANTS	

                                No. receiving                No. returning
Category                           samples                       data
Contractors
Industry
Foreign
Federal
State
Local
81
51
4
2
18
4
37
35
2
2
14
4
Total                                160                          94
     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^) requires that the dry gas meter be calibrated within
an accuracy of 2 percent.  Figure 2 shows that 47 percent of the reporting
laboratories attained this accuracy.  These results are not as good as
those from 1984.  Table 9 is a summary of previous audits that, like
Figure 2, shows a decline from 1984.

-------
                TABLE 9.  PREVIOUS RESULTS OF METHOD 5 AUDIT

Audit
0382
0982
0383
0983
0684
0685
Number of
analyses
827
769
763
614
631
633
Mean Median
(% from EPA
7.6
12.5
5.7
4.1
3.1
10.6
value)
2.5
2.5
2.2
1.9
1.9
2.2
Std. dev.
39.5
81.4
32.7
21.3
4.9
102.1
     The histogram in Figure 3 depicts the individual  results from all
participants of the 1985 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 71 percent of the standard deviations for
each set were within 0.3 percent.  Four percent of the 1985 data were identi-
fied as outliers using Chauvenet's Criterion.(5)  Before the outliers were
removed, the mean value (absolute) differed by 10.6 percent from the expected
value.  After deletion of outliers, this value was reduced to 2.3 percent.

-------
                                     LABORATORIES ACHIEVING SPECIFIED ACCURACY, percent
                                                 N>
                                                 O
                               U
                               O
                                    (II
                                    O
O)
o
Nj
o
00
o
(O
o
o
o
                        (Q
                        C
                        -^
                        CD
                                                                   3
                                                                   II
                                                                   O)
                                                                   w
                                                                   w
                        o
                        c
                        3
                        c_
                        D)
       31
       m
       Z
       O
       m
O
                        CD

                        0)
                        O
                        O
                        C

                        3
                        o
                        •a
                        to
                        o
                        •6'
                        Q>
                        3
                        W
                        3
                        CO
       m

       CO
       c
       3)
       m

       m
       Z

       CO
       Tl
                              m
                              •o
           a>
o
a.
01
0)
c
a.
rn

•o
CD
                                   00
                        O
                        en
                        oo
                        en
           (O

-------
   55
   50
CO
ill





I"
oc
o
CO
   40
   35
        DEGREE OF ACCURACY = 2%
         0382/827   0982/769   0383/763    0983/614   0684/631   0685/633




                             AUDIT / NUMBER OF SAMPLES
                         Figure 2. Previous results of Method 5 audit.
                                       11

-------
   20
   15
V
UJ
c  10
oc
o
CD
       NUMBER OF VOLUMES = 633
                MEDIAN = -1.0
       OUTLIERS
       REMOVED
                                        Mean = 0.7
                                                     OUTLIERS
                                                     REMOVED
<-12  -10   -8-6-4-20246

              DIFFERENCE FROM EPA VALUE, percent
                                                         8   10  >10
                    Figure 3. Results of method 5 audit, 0685.

-------
                                 SECTION 5
                               METHOD 6 AUDIT

     The Method 6 audit checks participants' ability to quantitatively analyze
Method 6 samples for sulfur content.  The audit set consists of five aqueous
dilutions of 10 N sulfuric acid in 25 milliliters (ml) sealed glass ampoules.
The analyst withdraws 5 ml from each ampoule, adds 30 ml of 3 percent 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 percent
isopropanol and thorin indicator are added, and the sample is titrated with
barium perchlorate 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 10 shows the categories of the participants and compares the
total number of participants requesting participation with the number return-
ing data.  In the 1985 audit, 66 percent of the 137 laboratories that received
the audit package returned data.
                   TABLE 10.  METHOD 6 AUDIT PARTICIPANTS
Category
No. receiving
   samples
No. returning
    data
Contractors
Industry
Foreign
Federal
State
Local
74
40
3
1
13
6
42
30
2
1
11
5
Total
     137
     91
     Table 11 shows the percentage of laboratories that achieved 2 percent
and 5 percent accuracy for each of the five different concentrations in the
Method 6 audit.   At least 62 and 67 percent of the reporting laboratories
achieved an accuracy within 5 percent for the two lower concentrations, but
83 percent of the laboratories showed an improvement on the higher concentra-
tions.  Five percent is used as the criterion because it was established as
the criterion for source SOg compliance audit samples.
                                     13

-------
               TABLE 11.  SUMMARY  OF SOURCE  S02 AUDITS—0585
Concentration                      ±2%                              ±5%
  (mg/DSCM)
198.3
221.2
701.6
1250.7
1326.9
28.1
34.1
58.9
39.6
64.0
61.8
67.0
83.3
87.9
86.5
                                         91
     Figure 4 shows the mean values for each S0£ concentration and the
confidence intervals (CI).   Each CI is calculated using the 95 percent
confidence level.  The CI for the lowest concentration is  ± 22.5 percent;
whereas the CI for the highest concentration is  ± 8.7  percent.  Overall,
the CI's for the SOg were higher than expected for this method because of
the high standard deviations.
                                     14

-------
PARTICIPANTS'VALUES. mg/DSCM
198.3 221.2 701.6 1250.7 1326.9
EXPECTED VALUES, mg/DSCM
Figure 4. SC>2 expected values compared to participants' values.
* 00 N> 0
O O O C
0 0 O C
II I I I I I I I I I II

LLJ





II I I I I I I I I I I

-------
                                 SECTION 6


                               METHOD 7 AUDIT
     The Method 7 audit checks participants'  ability to quantitatively
analyze Method 7 samples for nitrate content.  The NOX audit set consists
of five aqueous dilutions of a potassium nitrate solution in 2b-mL glass
ampoules that are autoclaved after sealing so that bacteria that
might attack the nitrate are destroyed.  The analyst withdraws 5 ml of
solution from an ampoule, adds this with 25 ml of 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.U)  After this treatment is completed, the
absorbance is measured at 410 nanometers (nm) with a calibrated spectro-
photometer.  In calculating the concentrations present, the participant
assumes that 2000 ml of stack gas was sampled.

     Table 12 shows the total number of laboratories requesting participa-
tion and the number that returned data for the 1985 Method 7 audit.  Sixty-
two percent of the 117 laboratories receiving the audit package returned
data.
	TABLE 12.  METHOD 7 AUDIT PARTICIPANTS	

                                 No. receiving               No. returning
Category                            samples                      data


Contractors                           67                          36
Industry                              36                          25
Foreign                                2                           2
Federal                                1                           0
State                                  6                           6
Local                                  5                           4
Total                                117                          73
                                     16

-------
     The percentage of laboratories that achieved 5 and 10 percent accuracy
for each of the five concentrations is shown in Table 13.  Ten percent is
used as the criterion because it was established as the criterion for the
source NOX compliance audit samples.  Sixty percent of the reporting
laboratories achieved an accuracy within 10 percent on the lowest concentra-
tion, and 70 to 80 percent achieved an accuracy within 10 percent on the
highest concentrations.
               TABLE 13.  SUMMARY OF SOURCE NOX AUDIT—0485
Concentration
(mg/DSCM)
99.5
119.5
298.6
338.4
537.5
± 5%
(*)
41.7
49.3
54.8
51.4
53.4
± 10%
(%}
58.3
69.9
80.8
72.2
71.2
                                                 73
     Figure 5 shows the means for each NOX concentration and the CI's.   The
CI's are calculated the same tfs for SOg.   The CI for the lowest concentration
is ± 5.8 percent, whereas the CI for the highest is 5.0 percent.  Overall,
the CI's for the NOX were very low because of low standard deviations.
                                     17

-------
    600
5  500
O
(/>
O
N.
O)
Ul
D
(A
O

DC
   400
    300
    200
    100
-  E
              99.5
                                                      537.5
              119.5       298.6       338.4


               EXPECTED VALUES, mg/DSCM


Figure 5. NOX expected values compared to participants' values
                                    18

-------
                                 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 concentra-
tion.  The coal audit checks participants' ability to analyze coal samples
for sulfur, ash, moisture, and Btu content.

     The coal audit samples consisted of two samples each with 50 grams (g)
of 60-mesh coal but with different parameter levels.  The following American
Society for Testing and Materials (ASTM) procedures^6) were recommended, but
not required, for participants1 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)

        • ASTM D-2015 (Standard Test Method for Gross Calorific Value
                      of Solid Fuel  by the Adiabatic Bomb Method).

     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 both audits, 80 percent of the laboratories that received the audit
package returned data.  Seventy-three of the same laboratories participated
in both audits and returned data.  Table 14 shows the total number of
laboratories requesting participation and the number that returned data for
coal  audits 0385 and 0985.
	TABLE 14.  COAL AUDIT PARTICIPANTS	

                         No. requesting samples        No.  returning data
 Category                    0385     0985                0385     0985
Contractors
Industry
Federal
State
Local
Total
57
42
1
11
4
115
52
37
1
12
5
107
42
37
0
10
3
92
42
33
0
10
5
90
19

-------
     Tables 15 and 16 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 percent were chosen  as the reporting
criteria for each of the four parameters  (sulfur,  moisture, ash,  and gross
calorific content).

     In the 0385, 90 percent of the laboratories were able to analyze both
sulfur samples within 10 percent of the expected value.   In the 0985 audit,
76 percent of the laboratories were able  to  analyze  the sulfur content of
the lower concentration sample within 10  percent of  the expected  value;
however, 90 percent were able to analyze  the high level sample within 10
percent of the expected value.  Seventy percent were within the 10  percent
criterion for both moisture concentrations in the 0385  audit, whereas 89
percent in the 0985 audit achieved 10 percent for the higher  moisture
concentration, but only 55 percent of the laboratories  achieved 10  percent
on the lower concentration.  For the ash  analysis and Btu content,  87 to
100 percent of the reporting laboratories were  able  to  achieve an accuracy
within 10 percent for both sample concentrations.

     Comparing the 0385 audit to the 0985 audit, there  was improvement in
the latter audit except for sulfur and moisture parameters.   In the  0985,
the sulfur and moisture levels were considerably lower, therefore causing a
greater chance for error.
                                    20

-------
                    TABLE 15.  SOURCE COAL AUDIT—0385
Expected
value
1.
4.
3.
3.

11.
20.
11,
13,
62
97
86
93

36
32
414
054
Laboratories
No. of accurate
analyses* within 5%
(%)
(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)

(1)
(2)
(1)
(2)
(1)
(2)
(1)
(2)
111
107
110
106
111
104
112
105

111
105
110
104
108
103
109
104
Sulfur
76
78
53
60
Moisture
43
39
40
45
Asji
96
97
93
96
Gross Calorific
98
99
98
98
.6
.5
.6
.4
.2
.4
.2
.7

.4
.1
.6
.2
.1
.0
.2
.1
Laboratories
accurate
within 10%
(%)
93.
95.
88.
85.
70.
68.
72.
74.

97.
97.
99.
99.
99.
99.
99.
99.
7
3
2
8
3
3
3
3

3
1
1
0
1
0
1
0
*  Numbers in parentheses indicate first and second analyses.
                                     21

-------
                 TABLE 16.  SOURCE COAL AUDIT—0985
Expected
value
0.72
2.35
1.61
17.45

4.33
17.92
12,300
12,374
No. of
analyses
(1) 108
(2) 102
(1) 106
(2) 100
(1) 106
(2) 100
(1) 108
(2) 102

(1) 107
(2) 101
(1) 105
(2) 99
(1) 102
(2) 97
(1) 104
(2) 99
Laboratories
accurate
within 5%
Sulfur
46.3
50.0
62.3
65.0
Moisture
31.1
35.0
66.7
70.6
Ash
68.2
69.3
94.3
92.9
Gross Calorific
99.0
100
89.4
92.9
Laboratories
accurate
within 10%
77.8
75.5
87.7
93.0
51.9
57.0
88.0
90.2

86.9
86.1
95.2
96.0
100
100
93.3
92.9
Numbers in parentheses indicate first and second analyses,
                                  22

-------
                                 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., 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, April  1984.

3.   Streib, E.W., T.J. 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, January 1985.  38 pp.

4.   Streib, E.W., T.J. Logan, and M.R. Midgett.   A Summary of the 1984
     EPA National Performance Audit Program on Source Measurements.   EPA-
     600/4-86-005, U.S. Environmental Protection  Agency, Research Triangle
     Park, North Carolina, February 1986.   34 pp.

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
     Standards—1984.  Volume 5.05. 01-050584-13, American Society for
     Testing and Materials, Philadelphia,  Pennsylvania.
                                    23

-------
       APPENDIX A



FREQUENCY DISTRIBUTIONS
            24

-------
en
                                     TABLE  17.   NATIONAL ORSAT  AUDIT  FREQUENCY  DISTRIBUTION



                            OF ABSOLUTE PERCENT  DIFFERENCES  OF EXPECTED  AND REPORTED VALUES—STUDY 0785
C02
Sample
no. No.
5000 91
Min. 10% 20% 30% 40% 50% 60% 70% 80% 90% Max. Mean Std.de v.
0.00 2.86 2.86 4.29 5.71 5.71 7.14 7.14 10.00 11.43 71.43 8.51 10.85
02
Sample
no. No.
5000 93
Min. 10% 20% 30% 40% 50% 60% 70% 80% 90% Max. Mean Std.dev.
0.00 0.00 0.00 0.83 0.83 1.67 1.67 1.67 3.33 4.17 51.67 4.24 10.44
CO
Sample
no. No.
Min. 10% 20% 30% 40% 50% 60% 70% 80% 90% Max. Mean Std.dev.
     5000     73    0.00   0.00   0.00   0.00   16.67   16.67   16.67   33.33    50.00    66.67    2950.00  102.06   480.43

-------
                              TABLE  18.   DGM  FREQUENCY  DISTRIBUTION OF ABSOLUTE PERCENT DIFFERENCE—0685
(S3
CT>
Sample
no.
633
630
617
609
600
591
588
Min.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10%
0.6
0.6
0.6
0.6
0.6
0.6
0.5
20%
1.0
1.0
1.0
1.0
1.0
1.0
0.9
30%
1.3
1.3
1.3
1.2
1.2
1.2
1.2
40%
1.8
1.7
1.7
1.7
1.7
1.7
1.7
50%
2.2
2.2
2.2
2.1
2.1
2.0
2.0
60%
2.6
2.6
2.5
2.5
2.5
2.4
2.4
70%
3.2
3.2
3.1
3.1
3.0
2.9
2.9
80%
4.1
4.1
4.0
3.9
3.7
3.6
3.6
90%
6.3
6.1
5.7
5.2
5.0
4.8
4.7
Max.
1,562.9
75.2
20.1
11.4
9.6
8.6
7.7
Mean
10.6
3.6
2.8
2.6
2.5
2.4
2.3
Std.dev.
102.1
6.5
2.6
2.1
1.9
1.7
1.7
Skewness
14.34
-0.00
-0.00
-0.00
-0.00
-0.01
-0.01
Median
2.2
2.2
2.2
2.2
2.1
2.0
2.0

-------
                    TABLE 19.   S02 FREQUENCY  DISTRIBUTION OF PERCENT DIFFERENCE—NO OUTLIERS REMOVED—AUDIT 0585
Sample no.
2
4
7
8
9
No.
90
89
89
91
91
Min.
0.00
0.00
0.01
0.00
0.00
10%
0.40
0.35
0.35
0.38
0.54
20%
0.77
1.01
0.60
0.73
0.99
30%
1.08
2.12
0.88
0.99
1.81
40%
1.51
2.67
1.12
1.22
2.31
50%
1.62
3.68
1.40
1.66
3.35
60%
2.15
4.69
1.80
1.97
3.48
70%
2.94
5.65
2.55
2.44
5.06
80%
4.22
7.31
3.93
3.17
6.69
90%
11.05
15.33
6.17
5.83
14.69
Max.
309.06
1186.89
304.61
296.58
301.67
Mean
8.18
26.47
9.93
9.38
13.72
Std.dev.
34.50
132.35
44.38
43.25
46.52
ro
                    TABLE 20.   NOX FREQUENCY  DISTRIBUTION OF PERCENT DIFFERENCE—NO OUTLIERS REMOVED—AUDIT 0485
Sample no.
1
3
5
6
8
No.
73
72
73
72
73
Min.
0.03
0.10
0.00
0.03
0.00
10%
0.37
1.51
0.65
0.62
0.50
20%
1.51
2.41
1.77
1.63
1.26
30%
2.31
2.61
2.75
2.63
2.09
40%
3.45
4.52
3.48
3.63
4.44
50%
4.39
6.63
4.52
4.43
4.60
60%
5.66
10.05
6.27
6.62
7.95
70%
7.23
14.87
9.08
9.04
9.62
80%
9.21
26.43
15.40
15.48
19.33
90%
20.73
33.97
32.76
27.45
32.13
Max.
79.60
95.18
98.81
117.55
79.33
Mean
9.23
15.01
11.69
11.61
11.58
Std.dev.
13.81
19.43
17.77
18.50
15.31

-------
                                        TABLE 21.  NATIONAL COAL AUDIT FREQUENCY DISTRIBUTION

                            OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES—STUDY 0385
ro
oo

Sample no.
1000
2000

Sample no.
1000
2000

Sample no.
1000
2000

Sample no.
1000
2000

No.
218
216

No.
217
215

No.
216
214

No.
213
211

Min.
0.00
0.00

Min.
0.00
0.00

Min.
0.00
0.00

Min.
0.00
0.00

10%
0.62
0.60

10%
0.76
1.04

10%
0.09
0.25

10%
0.08
0.09

20%
1.23
1.61

20%
1.27
2.07

20%
0.26
0.44

20%
0.18
0.16

30%
1.85
2.21

30%
3.05
3.37

30%
0.35
0.79

30%
0.25
0.25

40%
1.85
3.22

40%
4.58
4.66

40%
0.44
0.94

40%
0.36
0.33
Sulfur
50%
3.09
4.23
Moi
50%
6.62
6.22

50%
0.53 0
1.28 1
Gross
50%
0.47
0.46
60%
3.70
5.23
sture
60%
8.14
8.03
Ash
60%
.79
.37
Calori
60%
0.59
0.63

70%
4.32
6.84

70%
9.41
10.10

70%
1.06
1.87
fie
70%
0.78
0.77

80%
5.56
7.85

80%
11.45
15.28

80%
1.41
2.26

80%
1.03
1.05

90%
8.02
11.47

90%
22.39
22.54

90%
2.64
3.54

90%
1.55
1.63

Max.
66.67
42.25

Max.
417.05
89.12

Max.
24.12
77.36

Max.
17.59
19.98

Mean
4.34
5.76

Mean
13.01
10.05

Mean
1.47
2.32

Mean
0.92
0.92

Std.dev.
6.98
6.22

Std.dev.
40.70
12.30

Std.dev.
3.51
7.46

Std.dev.
1.86
2.07

-------
                                        TABLE  22.   NATIONAL  COAL  AUDIT  FREQUENCY  DISTRIBUTION

                             OF ABSOLUTE PERCENT DIFFERENCES OF EXPECTED AND REPORTED VALUES—STUDY 0985
ro
vo
Sulfur
Sample no.
3000
4000
No. Min.
210 0.00
206 0.00
10% 20%
1.39 1.39
0.43 0.85
30%
2.78
1.70
40%
4.17
2.98
50%
5.56
3.83
60%
6.94
4.26
70%
8.33
5.96
80%
11.11
8.09
90%
18.06
9.79
Max
2256.
55.
Mean
94 31.37
74 5.51
Std.dev.
219.04
7.60
Moisture
Sample no.
3000
4000
No. Min.
210 0.11
206 0.00
10% 20%
0.63 1.43
1.24 2.48
30%
2.12
4.35
40%
2.58
6.21
50%
3.15
8.70
60%
4.01
11.18
70%
5.27
14.91
80%
6.76
19.88
90%
10.09
32.30
Max. Mean
99
1009
.08 6.64
.94 25.77
Std.dev.
14.28
106.36
Ash
Sample no.
3000
4000
No. Min.
208 0.00
204 0.00
10% 20%
0.46 1.15
0.22 0.45
30%
1.62
0.61
40%
2.31
0.78
50%
3.00
1.17
60%
3.70
1.45
70%
5.08
2.01
80%
6.93
2.51
90%
14.09
4.13
Max.
83.83
87.67
Mean Std.
5.35 9.
3.82 12.
dev.
10
53
Gross Calorific
Sample no.
3000
4000
No. Min.
203 0.00
199 0.00
10% 20%
0.28 0.45
0.07 0.15
30%
0.61
0.24
40%
0.79
0.37
50%
0.99
0.50
60%
1.12
0.64
70%
1.48
0.89
80%
1.92
1.08
90%
3.51
1.63
Max.
18.52
9.34
Mean Std
2.30 4.
0.80 1.
.dev.
19
04

-------
             APPENDIX B



INSTRUCTIONS FOR EPA AUDIT MATERIALS
                 30

-------
                                                                Page 1  of 2
              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 C09, 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
                                     31

-------
                                              Page 2 of 2
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
           inn
              o
              o
 NJ
o
                          32

-------
                                                                  Page 1 of 2
            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.
                                      33

-------
                                                                  Page 2 of 2
PROCEDURE:   (continued)
          Make triplicate quality assurance runs.   For 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 m3ft3 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.
                                     34

-------
     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. j*2,  No.  160, Part II, Thursday,
       August 18,  1977,  PP  1*1782-U178U.   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, Cso2 (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.

U.   Record the reference standard sample numbers and their corresponding
     SOg 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.
                                   35

-------
   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 jjg N02 per sample using Equation 7-3.  Calculate
    the sample concentration, C (concentration of  NOX  as NC>2» 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.
                                     3G

-------
                    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
     1n weight percent with gross calorific  value  reported  as  BTU/lb.

 3.   All methods used In the analysis of these coal samples should follow
     American Society for Testing and Materials  (ASTM)  recomnended 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  1s
     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. Strelb
                    Quality Assurance Division (MD-77)
                    Environmental Monitoring Systems Laboratory
                    U.S. Environmental  Protection Agency
                    Research Triangle Park, NC  27711

     A postpaid return envelope Is enclosed for this purpose.

9.   If you have any questions concerning this or any source nethod audit,
     please call (919/541-7834).

                                     ?7
                                     0'         *U.S.Government Printing Office: 1987 - 748-121/40689

-------