United States
Environmental Protection
Agency
industrial Environmental Research EPA 600 4-80-017
Laboratory June 1980
Research Triangle Park NC 27711
Research and Development
Summary of Audit
Performance
Measurement of S02,
N(X Sulfate, Nitrate,
Lead, Hi-Vol Flow
Rate—1978
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
SUMMARY OF AUDIT PERFORMANCE: MEASUREMENT OF
S02, N02, CO, SULFATE, NITRATE, LEAD, AND HI-VOL FLOW RATE
1978
Steven M. Bromberg, Robert L. Lampe, and Berne I. Bennett
Quality Assurance Division
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
QUALITY ASSURANCE DIVISION
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protections-Agency, and approved for publica-
tion. Mention of trade names or commercial products does not constitute
endorsement 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 ecology, to provide innovative means of monitoring compliance with
regulations, and to evaluate the effectiveness of health and environmental
protection efforts through the monitoring of long-term trends. The Environ-
mental Monitoring Systems Laboratory, Research Triangle Park, North Carolina,
has responsibility for: assessment of environmental monitoring technology
and systems; implementation of agency-wide quality assurance programs for
air pollution measurement systems; and supplying technical support to other
groups in the Agency including the Office of Air, Noise and Radiation, the
Office of Toxic Substances and the Office of Enforcement.
The following investigation was conducted as part of the routine
Environmental Protection Agency quality assurance program. Results of
quality control audits sponsored by the Environmental Monitoring Systems
Laboratory for calendar year 1978 are presented. Measurement methods for
S02, N02, CO, Pb, S0»=, NOZ, and hi-vol flow rate were audited. Preceding
reports in this series have dealt with similar topics for 1976 and 1977.
Thomas R. Hauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27711
m
-------
ABSTRACT
The Quality Assurance Division of the Environmental Monitoring Systems
Laboratory, Research Triangle Park, North Carolina conducts an ambient air
audit program. Measurement principles for sulfur dioxide, nitrogen dioxide,
carbon monoxide, sulfate-nitrate, and lead are audited on a semiannual basis.
Blind samples, the concentrations of which are known only to the EPA, are
sent to the participating laboratories. Hi-vol sampler flow rate audits
are conducted annually using a modified orifice. The analytical results
are returned to the Quality Assurance Division for evaluation. After
processing, an individual report is returned to each participant.
This report contains a summary of audit results obtained during
calendar year 1978.
iv
-------
CONTENTS
Abstract iii
Figures , v
Tables vi
Acknowledgment v'i'ii
1. Introduction 1
2. Program Coordination 3
3. Statistical Approach 5
4. Audit Materials
S02 8
N02 9
CO 9
SOj-NOg 11
Pb 11
Hi-Vol Reference Flow Device 13
5. Results
S02 16
N02 27
CO 38
SOj 48
NOg 59
Pb 6f
Hi-Vol Flow Rate 77
References 83
Appendices 84
-------
FIGURES
Number Page
1 Example of individual report 4
2 ReF mounted on hi-vol sampler 14
3 ReF with resistance plate 14
4 Plot of S02 results 18
5 Plot of S02 absolute percent differences 21
6 Plot of S02 slope and intercept . 24
7 Plot of N02 results 30
8 Plot of N02 absolute percent differences 32
9 Plot of N02 slope and intercept 35
10 Plot of CO results . . 41
11 Plot of CO absolute percent differences 43
12 Plot of CO slope and intercept 46
13 Plot of SOj results 51
14 Plot of SO^ absolute percent differences 53
15 Plot of S0]j slope and intercept 54
16 Plot of NOg results 61
17 Plot of NOZ absolute percent differences 63
18 Plot of NOZ slope and intercept 66
19 Plot of Pb absolute percent differences 73
20 Hi-vol flow rate histogram - 1978 80
vi
-------
TABLES
Number Page
1 Mean and Standard Deviation of S02 Acceptance Analysis lo
2 Mean and Standard Deviation of N02 Acceptance Analysis 10
3 Mean and Standard Deviation of CO Acceptance Analysis 12
4 Sulfate and Nitrate Sample Values 12
5 Lead Sample Values 13
6 SCL Agency Distribution 16
7 S02 Analytical Methods 17
8 S02 Sample and Target Ranges 18
9 S02 Absolute Percent Difference 20
10 S02 Survey Statistics 22
11 Summary of S02 Analytical Methods Used by Outlier Labs 25
12 Mean and Standard Deviation of S0? Results by Analytical
Method 26
13 N02 Agency Distribution 28
14 N02 Analytical Methods 28
15 NO,, Samnle and Target Ranges 29
16 N02 Absolute Percent Difference 32
17 N02 Survey Statistics 34
18 Summary of N02 Analytical Methods Used by Outlier Labs 36
19 Mean and Standard Deviation of NO,, Results by Analytical
Method 7 37
20 CO Agency Distribution 39
21 CO Analytical Methods 39
22 CO Sample and Target Ranges 40
23 CO Absolute Percent Difference 43
24 CO Survey Statistics 45
25 Summary of CO Analytical Methods Used by Outlier Instruments . . 47
vn
-------
Number Page
26 Mean and Standard Deviation of CO Results by Analytical
Method 47
27 S(L Agency Distribution 49
28 SO^ Analytical Methods 49
29 SO^ Sample and Target Ranges 50
30 S0| Absolute Percent Difference 52
31 SO! Summary Statistics 55
32 Summary of SO^ Analytical Methods Used by Outlier Labs ... 55
33 Mean and Standard Deviation of S0« Results by Analytical
Method 57
34 NOZ Agency Distribution 59
35 NOg Analytical Methods 60
36 NOo Sample and Target Ranges 60
37 NO, Absolute Percent Difference 62
38 NO, Summary Statistics • • • 65
39 Summary of NOl Analytical Methods Used by Outlier Labs ... 67
40 Mean and Standard Deviation of NO" Results by Analytical
Method : 68
41 Pb Agency Distribution 69
42 Pb Analytical Methods ., 70
43 Pb Extraction Procedures 70
44 Pb Sample and Target Ranges 71
45 Pb Absolute Percent Difference 72
46 Pb Summary Statistics 74
47 Summary of Pb Analytical Methods Used by Outlier Labs ... 75
48 Hi-Vol Flow Rate Agency Distribution 77
49 Hi-Vol Flow Rate Measurement Methods 77
50 Hi-Vol Flow Rate Percent Differences 79
51 Summary of Hi-Vol Flow Rate Measurement Methods si
viii
-------
ACKNOWLEDGEMENTS
The authors would like to thank all the participants for their coopera-
tion during the past year. Also due a word of thanks are the programmers
of the Statistical and Technical Analysis Branch, EMSL, Research Triangle
Park, North Carolina, for developing the data management systems necessary
to store and summarize the audit data; the EPA Regional Quality Control
Coordinators; and the technical reviewers for reviewing a very tedious
document.
IX
-------
SECTION 1
INTRODUCTION
Since 1972 the Environmental Protection Agency (EPA) has been engaged in
a performance audit program of various monitoring groups throughout the
United States and in several foreign countries. The audit program is only
one part of an overall quality assurance program, therefore, the results
should not be construed as an absolute indicator of data quality. Used along
with information obtained from an internal quality control program, however,
the conclusions can be quite meaningful.
The purposes of the audit program are twofold. The first, from a
participant standpoint, is most important. Agencies are furnished a means of
rapid self-evaluation of the specific operation under study. The second
objective of the program is to provide EPA with a continuing index of the
validity of data reported to air quality data banks.
The program is being coordinated through the 10 EPA Regional Offices
(RO) by the Quality Assurance Division (QAD) of the Environmental Monitoring
Systems Laboratory (EMSL), Environmental Research Center, Research Triangle
Park, North Carolina 27711. Comments, questions, or applications to parti-
cipate in the program should be sent to the above address.
This publication is the third of a continuing series of yearly summary
reports. The document covers the period from January through December, 1978.
Users of the information contained in this report should take note
1
-------
of some limitations imposed in gathering the data. With the exception of the
carbon monoxide (CO) audit, the surveys checked only a portion of the entire
system. The sulfur dioxide (SO,,), nitrogen dioxide (NOp), sulfate-nitrate
(SOl-NO^), and lead (Pb) surveys examined only the analytical portion of the
system. The Hi-vol audit checked only the flow rate portion of the method.
No restrictions were placed on the methodology used by the participants
(however, the method used had to be compatible with the audit samples).
To the extent possible, the various methods employed were documented.
The following sections include discussions of the program operation,
descriptions of the audit materials, the statistical approach used to analyze
the data, and the results. The Results Section includes data for calendar
year, 1978. The appendix contains the raw data from the audit sample analysis,
-------
SECTION 2
PROGRAM COORDINATION
Participants in the audits were selected by the Regional Quality Control
Coordinator in each of the 10 Regions. Once a potential participant has
received audit samples for a particular oollutant, he is automatically
notified of subsequent surveys for that pollutant. Participants are assigned
an identification number which remains with the agency throughout all audits
for all pollutants.
After the audit roster is completed for a particular survey, instruc-
tional materials and unknown samples are mailed. The participants are
allowed 5 to 6 weeks in which to return their results. After these results
are entered into a data bank, individual reports are returned to the parti-
cipants. This report indicates the acceptable ranges for each samele as well
as the value reported by the agency. Figure 1 is an example of an individual
report for SO!.
In order to determine whether incorrect results are due to inherent
problems with a laboratory procedure or to just a "single event", recheck
samples are sent to laboratories whose results do not meet certain criteria.
Thus, by having a second chance to analyze a set of samples, real deficiencies
can be distinguished from one-time problems.
-------
123456
Mr.' J. Ooe
5432 Anywhere St.
Vourtown, USA
543ELE
3113
l| I I
5251
6116
911?
IMTFR-l.*nOR*TORy STi'CW RF5ULTS
( AUGUST 1977 I
POLLUTANT - SflH
BU'QBIEO VALUE'
. ISO
S»UELE_BftUGE
.000
9.900
I I.810
11.090
2.750
1 .760
10.515
17.825
15.390
?~.H50
_ .000
11.655
I N. 1 75
17.010
.1.150
7«?15
9.1J5
I I .ITS
1_J .770
2.550
1«7"5
.nip
I 7.7*5
3.1HP
7.115
Iq>80M !
| q»917
0 10*175!
•» !
T !
r. s
! .X
! X
.Mfl2 5.3Q1 10.175 IH.917
EP« RTSUITS
T • .lAflx »O.I?H
Figure 1. Example of Individual Report.
-------
SECTION 3
STATISTICAL APPROACH
A primary concern at the onset of the program was the establishment of
acceptance criteria to enable individual participants to judge their per-
formance. Initially, there was little guidance, and the choices of accep-
table limits were somewhat arbitrary. As more data was gathered through the
audit program, subsequent and more realistic acceptance limits were set.
For convenience, two acceptable ranges were defined. The "Sample range"
contains the variability attributable to samole material. This estimation
was based on the repeated analysis of several samples from each concentration
range by one laboratory. Using the precision of the set of samples under
consideration plus past data from similar samples, the Sample ranges were
chosen. These ranges should be the goal of each agency; it is unreasonable
to expect all laboratories to fall within these ranges on any given analysis.
However, falling within these values repeatedly indicates a facility with
excellent precision and accuracy. Falling outside the Sample ranges does
not necessarily indicate problems.
The "Target range" has been constructed to include sample variability
and interlaboratory variability. A laboratory falling outside the Sample
range but within the Target range should feel comfortable with its results.
However, if their results fall outside the Target range, the accuracy of the
analysis should be considered suspect. The Target range is based on the past
-------
performance of all participants. The percent difference between reported
results and the true value (determined by QAD) was established for past
surveys. Applying this percent difference to all samples over all surveys,
an average difference was determined which was used to calculate the Target
ranges.
When evaluating results based on the acceptable ranges, one must con-
sider the usage of the data. For some purposes the Sample range may be too
wide; for others the Target range may be too narrow. Thus, judgment should
be exercised whenever comparing ranges with results.
A preferable method of evaluating performance "after the fact" is to
compare individual results with the tables entitled "Absolute Percent Differ-
ence" (See Results Section). These tables are frequency distribution tables
of the percent difference between EPA and reported values. The relationship
of individual performance to overall study performance can be determined
using these tables.
Beginning with calendar year 1979, the use of Sample and Target ranges
will end. Replacing these QAD estimates of performance will be frequency
distribution tables based on past performances. This new reporting format
will allow each monitoring group to judge not only their accuracy, but also
their relative performance to other groups doing similar analyses.
Two outlier tests were used as part of the overall analysis and to
screen data for further analysis. The first test was employed to screen
results that were grossly in error. To be eliminated from further analyses,
a laboratory had to report all samples outside the Target ranges for the
respective pollutants. These data appear only in the Appendix and are
marked with an asterisk (*). No further statistical analyses in this report
6
-------
contain these data. Data aonearina in table columns labeled "All data", or
containing no specific designation, are results subjected to the first test.
A second outlier test, using Chauvenet's technique (1) was also used.
Depending upon the number of results for each concentration, a factor ranging
from 2.1 for 20 samples to 3.3 for 300 samples, was chosen. This factor was
multiplied by the standard deviation of the sample as determined from the
participant results. Results outside the range determined by the expression
([factor] x [standard deviation]) ± study mean were identified as outliers.
These data are identified in the "Outliers removed" columns of selected
tables.
Several summaries have been used to condense the large amounts of data
into a more manageable form. Tables in the Appendix are sequential listings
of all data by sample concentration for each audit. Statistical summaries of
the sequentially listed data are also presented.
-------
SECTION 4
AUDIT MATERIALS
AMBIENT SULFUR DIOXIDE SAMPLES
The commercially produced sample material consisted of freeze-dried
mixtures of sodium sulfite and potassium tetrachloromercurate (KM) contained
in 5 ml sealed glass ampoules. Sample sets were comprised of 5 ampoules
containing approximately 3 to 61 ug of S02 equivalent per container. The
samples were immediately placed in freezers upon receipt with the expectation
that low temperatures would nreserve the integrity of the material. Initial
EPA analyses were performed immediately after receipt. Reanalyses after
several months demonstrated that freezing did not completely stabilize the
sulfite content as indicated by the continued decline in S0? levels.
Subsequent analyses of the samples have shown that, while the decay is not
completely eliminated, the rate has been substantially reduced. Analyses
were performed by the reference method for the determination of S0? in the
atmosphere (pararosaniline method)(2). The sample, when dissolved in 0.04 N
TCM forms a dichlorosulfitomercurate complex. This complex is reacted with
pararosaniline and formaldehyde to form intensely colored pararosaniline
sulfonic acid. The absorbance of the solution is measured spectrophoto-
metrically at 548 nm.
Ten samples from each concentration were analyzed (Table 1) and are
used in determining the acceptable ranges reported to participants. The
tabulated values are based on the assumption that individual samples were
8
-------
collected in 50 ml of absorbing reagent with a total sample air volume of
300£.
AMBIENT NITROGEN DIOXIDE SAMPLES
The commercially produced samples consisted of 4 ml of aqueous sodium
nitrite (NaNO,,) in 5 ml glass vials with inert screw-cap closures. A set
consisted of five vials. When mixed with absorbing reagents, the samples
simulated ambient samples ranging from 0.12 to 0.93 yg/ml.
EPA analysis of 10 samples from each concentration was performed using
an equivalent method for the determination of N02 in ambient air (3).
Measurements were made on a Varian Model 635 UV-Vis spectrophotometer at
540 nm. Aqueous calibration standards were prepared to encompass a linear
range from 0 to 1.6 yg/ml of N02 (Table 2). The values contained in Table 2
are utilized in determining the acceptable ranges reported to participants.
Values are based on the assumption that the sample was collected in 50 ml of
absorbing reagent.
AMBIENT CARBON MONOXIDE SAMPLES
Samples consisted of commercially produced mixtures of CO and artificial
air. Mixtures also contained methane (CH») and 350 ppm of carbon dioxide
(C0?). Specially treated aluminum cylinders were utilized to improve gas
stability. Sample concentrations ranged from approximately 4 to 42 ppm of
CO. Each participant received a set of 3 cylinders, one from each concen-
tration level.
Analysis by EPA of cylinders from each concentration was performed using
a Bendix Model 8501 NDIR analyzer. National Bureau of Standards, Standard
Reference material (SRM) gases were used as reference standards. Three SRM's
-------
TABLE 1. MEAN AND STANDARD DEVIATION OF S02 ACCEPTANCE ANALYSIS (yg/m3)
S09 audit
^0478
S0« audit
^1078
Cone. 1
Mean Std. dev
21.70 0.70
9.07 1.17
Cone. 2
Mean Std. dev
64.41 1.10
38.80 2.27
Cone. 3
Mean Std. dev.
64.72 1.10
88.90 3.77
Cone.
Mean Std
126.20 1
128.30 3
4 Cone. 5
. dev. Mean Std. dev.
.70 204.00 3.20
.70 192.20 3.17
TABLE 2.
MEAN AND STANDARD
DEVIATION OF N02 ACCEPTANCE ANALYSIS (yg/ml)
N09 audit
L 0678
N09 audit
* 1278
Cone. 1
Mean Std. dev
0.12 0.004
0.26 0.007
Cone. 2
Mean Std. dev
0.24 0.004
0.39 0.004
Cone. 3
Mean Std. dev.
0.24 0.003
0.52 0.004
Cone.
Mean Std
0.37 0.
0.70 0.
4 Cone. 5
. dev. Mean Std. dev.
004 0.69 0.004
005 0.93 0.008
-------
at nominal concentrations of 9, 46, and 95 ppm CO were used as the primary
reference gases.
Ten samples from each concentration were analyzed. Table 3 lists the
results of these analyses. These values were later used in determining the
acceptable ranges reported to participants.
SULFATE-NITRATE SAMPLES
The commercially prepared samples consisted of 19 mm by 20 cm (0.75 x
8 in) fiberglass filter strips with depositions of potassium sulfate (K2SO.)
and lead nitrate (PbCNO-Jo)- Filter strip sample sets were comprised of
combinations of differing SO^ and NO, concentrations. Each strip was packaged
in a plastic envelope. The concentration of sulfate ranged from 0 (blank) to
•3 O
40 yg/m . Nitrate concentrations ranged from 0 (blank) to 14 yg/m . Con-
centrations were calculated assuming that samples were collected on a 20 by
25.4 cm (8 x 10 in) filter with a total air volume of 2000 m .
It was felt that gravimetric preparation of the solution deposited onto
the filter strips was more accurate than an analysis using existing pro-
cedures. Thus, the values accepted as "true values" are the vendor certifi-
cations. Verification analyses ensured that the accuracy and precision of
the samples were acceptable. Table 4 lists the concentrations of samples
used during the audit.
LEAD SAMPLES
The commercially prepared samples consisted of 19 mm by 20 cm (0.75 x
8 in) fiberglass filter strips with depositions of lead nitrate (Pb[NO-]?).
Filter strip sample sets were comprised of combinations of differing lead
concentrations; each packaged in a plastic envelope. The concentrations of
11
-------
TABLE 3. MEAN AND STANDARD DEVIATION OF CO ACCEPTANCE ANALYSIS (PPM)
ro
CO audit - 0378
CO audit - 0978
Cone
Mean
7.15
3.82
. 1
Std. dev.
0.10
0.05
Cone. 2
Mean Std. dev
20.
14.
23 0.09
56 0.18
Cone. 3
Mean Std. dev.
42.08 0.08
36.45 0.07
TABLE 4
. SULFATE
AND NITRATE
SAMPLE VALUES
(yg/m3)
Cone. 1
S04 N03
SCVN03 0 00 0 00
audit 0278 '
Cone
so4
1.80
. 2
N03
0.45
Cone. 3
S04 N03
2.70 1.44
Cone. 4
S04 NO
9.60 10.
Cone. 5 Cone.
3 S04 N03 S04
50 12.00 12.00 14.40
6
N03
14.40
0.72 3.00 1.80 11.10 4.80 13.50 6.00 26.40 9.60 39.60 11.40
-------
o
lead ranged from 0 (blank) to approximately 13 yg/m . Concentrations were
calculated assuming the samples were collected on a 20 by 25.4 cm (8 x 10 in)
filter with a total air volume of 2000 m .
Gravimetric preparation of the solution deposited onto the filter strips
was assumed more accurate than analysis using existing procedures. Thus,
"true values" were the vendor certifications. Verification analyses ensured
that the accuracy and precision of the samples were acceptable. Table 5
lists the concentrations of the samples used during the audit.
TABLE 5. LEAD SAMPLE VALUES (yg/m3)
Cone. 1 Cone. ?. Cone. 3 Cone., 4 Cone. 5 Cone. 6
0.00 0.60 2.04 6.60 10.20 12.60
0.00 0.42 1.50 4.95 9.90 12.00
HI-VOL REFERENCE FLOW DEVICE (ReF)
A single ReF was supplied to each participating agency. Organizations
were instructed to check as many Hi-vol sampling units as feasible within
the allotted time. The auditing unit received by each laboratory consisted
of a modified orifice (ReF), wind deflector, manometer, and resistance
plates (to change flow rates).
During auditing of the flow rate of a Hi-vol sampler, the ReF was
mounted on top of the sampler replacing the filter face plate (Figure 2).
A wind deflector was necessary to prevent fluctuation in the readings due to
wind flow across the orifice. The resistance plates, when inserted into the
ReF, simulated various filter loading conditions (Figure 3).
13
-------
Figure 2. ReF mounted on hi-vol sampler.
Figure 3. ReF with resistance plate.
14
-------
By calibrating each ReF with a positive displacement meter (Roots
meter), in conjunction with measurements of pressure drops and temperatures,
an individual calibration curve in the form of an orifice equation was derived.
The equation shown below was used to determine the "K" orifice constant for
each unit.
Q1 = AYC /APTl (1)
where Q-j = volumetric flow at conditions of T, and P-, (m /min)
2
A = area of orifice (in )
Y = expansion factor
C = orifice coefficient
AP = pressure drop across orifice (in H?0)
PI = upstream pressure (barometric pressure, mm Hg)
T-| = upstream temperature (ambient temperature, °K)
Because A is constant for a given orifice, and Y and C are essentially
constant over the flow range in question, a new orifice constant "K" was
defined as:
K = AYC (2)
Thus, the orifice equation becomes:
•••••••••••••••••••••IB
» n T
(3)
During calibration of the ReF, Q1, AP, T-j, and P-| were also measured. The
constant K was determined by regressing a series of 0-| measurements onto
the square root of the values under the radical.
During an audit, field personnel measured AP, T^, and P,. By
knowing K, the "true flow" can be calculated. This flow was comnared with
the flow rate measured by the Hi-vol sensor to determine the accuracy of flow
measurements.
15
-------
SECTION 5
RESULTS
AMBIENT SULFUR DIOXIDE
Parti cipant Ch aracteristics
Sulfur dioxide study number 0478 began in April, 1978. Out of 178
sample sets requested by participants, 125 sets of data were returned for a
response rate of 70 percent. Study number 1078 began in October, 1978. Out
of 131 sample sets requested, 100 sets of data were returned for a response
rate of 76 percent. The total number of laboratories discontinuing SOp wet
chemical analysis in favor of continuous monitors continues to increase.
This is supported by the steady decrease in participants.
Table 6 indicates the monitoring agency distribution.
TABLE 6. S02 -AGENCY DISTRIBUTION
Foreign EPA State Local Private Total
S02 audit - 0478
Agencies
requesting samples
Agencies
returning data
3
1
6
3
62
51
74
51
33
19
178
125
S02 audit - 1078
Agencies
requesting samnles 1
Agencies
returning data 1
49
37
56
49
24
14
131
100
16
-------
Procedures for analyses were grouped into three broad categories:
manual pararosaniline, automated pararosaniline, and other. Table 7 lists
the analytical methods employed and the number of respondents using a
particular method.
TABLE 7. S02 ANALYTICAL METHODS
Method _ Agencies using method
Audit 0478 Audit 1078
Manual pararosaniline 93 79
Automated pararosaniline 29 21
Other 3 0
It should be noted that laboratories tend to define the procedure used in
very general terms. Thus, an agency using the manual pararosaniline technique
may not have used the method exactly as it appeared in 40 CFR 50.11 (2).
Acceptable Ranges
As previously described, two ranges were used as one means of judging
performance. The Sample and Target ranges for the studies are listed in
Table 8 and apply to sample values in increasing concentrations.
Sample Ranges were not determined by a method described earlier in our
series (4) and were arbitrarily set at one half the Target ranges. It was
determined from earlier studies that the average percent difference between
the reported results and EPA determined values was ±20 percent. Thus, this
value has been used for all Target ranges.
17
-------
TABLE 8. S02 SAMPLE AND TARGET RANGES (35)
Sample
Target
Cone. 1
range ±10
range ±20
Cone. 2
±10
±20
Cone. 3
±10
±20
Cone. 4
±10
±20
Cone. 5
±10
±20
Using those criteria, a tabulation was made of the number of agencies
reporting results within the ranges. For audit 0478, 48 (38%) of the
agencies reported all 5 results within the Target ranges, while 12 (10%) of
the laboratories reported all results outside these ranges. The corresoonding
figures for audit 1078 showed 6 (6%) and 8 (8%), respectively. These last
figures were considered suspect because of apparent problems with the audit
samples. Figure 4 shows a running tally of the above .values.
60'
co
LU
i—i
O
z
LLJ
CD
50*
40.
30
20
10
ALL 5 SAMPLES IN
TARGET RANGES
ALL 5 SAMPLES OUTSIDE
TARGET RANGES
r-.
if)
O
U3
!•-•
i^
«d-
O
00
i-^
*d-
O
00
r^
o
STUDY NUMBER
Figure 4. Plot of S02 Results (five samples within and outside Target Ranges)
18
-------
Data Summary
After using Target ranges as one means of eliminating nonrepresentative
data, it was decided that laboratories reporting all 5 results outside the
Target ranges (i.e., results greater than +_2Q% of the EPA values) would be
removed from the data base and excluded from further evaluations. The
Target ranges are broad enough that, unless the determination is totally out
of control, at least one value should fall within the ranges if the labora-
tory is performing adequately in comparison to most of the study population.
If the agency was not performing well, its data was excluded from the
summaries.
This discussion will be concerned with the reported results minus the
data sets meeting the exclusion criterion. These data points are marked in
the Appendix by an asterisk (*).
Table 9 is a frequency distribution of the percent difference between
the reported and EPA values for each sample concentration. The differences
were calculated by the following equation:
absolute percent difference =
reported value - EPA value
EPA value
x 100 (4)
A frequency distribution was then constructed and appears below. For
example, for audit 0478, 50 percent of the reported results for sample
concentration 1 were less than or equal to an absolute percent difference of
15.2 percent. Note that the "All samples" line is not an average of the
values appearing above it, but is the resulting distribution when all data
is compiled, regardless of concentration.
The data in Table 9 are very useful for laboratories trying to determine
their relative performance. For example, for audit 0478, only 10 percent of
19
-------
the labs reporting for concentration 1 had a percent difference of 2.3 percent
or less, while 50 percent of the laboratories reported a percent difference
of 15.2 percent or less. Table 9 also indicates the average percent difference
for all laboratories for all samples. These are presented graphically in
Fiqure 5 along with corresponding values from previous audits.
TABLE 9. S02 ABSOLUTE PERCENT DIFFERENCE
Concentration
so2
Al
so2
Al
audit - 0478
1
2
3
4
5
1 samples
audit - 1078
1
2
3
4
5
1 samples
No.
107
113
112
112
112
556
86
92
91
90
91
450
Min.
0.0
0.0
0.1
0.3
0.0
0.0
3.6
0.1
0.2
0.1
0.8
0.1
10%
2.3
0.6
0.9
1.7
1.1
1.0
11.5
1.2
1.3
1.4
4.3
2.0
30%
7.5
3.3
2.7
3.0
2.8
3.4
39.2
5.5
4.4
3.1
7.9
6.0
50%
15.2
6.4
5.9
5.2
5.8
6.6
55.1
10.5
8.9
5.0
11.9
11.7
70%
29.0
10.5
10.3
7.8
10.0
11.7
73.3
18.9
13.4
7.9
16.0
18.9
90%
49.3
18.6
21.5
16.8
24.2
30.9
152.2
44.2
18.4
16.8
22.3
58.5
Max.
1022.
60.
68.
58.
89.
1022.
374.
81.
54.
38.
42.
374.
9
5
6
5
2
9
1
6
6
7
6
1
Mean
30.9
9.5
10.3
8.7
11.6
14.0
71.3
18.4
10.9
7.6
12.9
23.7
All data received (except omitted data) were grouped according to
concentration. The results for each sample value are listed in the Appendix
in increasing concentration. Summary statistics which appear in Table 10
and at the top of each listing in the Appendix do not include outliers.
Table 10 tabulates the summary statistics based on the reported data.
The "Outliers removed" columns contain data on which the outlier criterion
was applied; all data meeting the second criterion (see Statistical Approach
20
-------
Section).were removed. Note that the variation in the number of samples
fron concentration to concentration is due to laboratory accidents and
damaged samples.
60,
50
LU
O
S 40J
u_
Ll-
I—i
!: so j
O
oo
CQ
20-
10 •
in
r->
in
o
10
CO
r-.
O i—
STUDY NUMBER
00
o
O r—
Figure 5. Plot of S02 absolute percent differences
An examination of Tables 9 and 10 reveal that there were no apparent
problems with Audit 0478. Except for concentration 1, 50 percent of the
laboratories had percent differences of less than 10 percent. Accuracy and
skewness values also indicate a normal distribution and acceptable accuracies.
21
-------
TABLE 10. S02 SURVEY STATISTICS (yg/m3)
Concentration 1
S02 Audit - 0478
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
S02 audit - 1078
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
All
data
107
21.7
20.7
19.5
242.0
22.5
108.4
9.1
-10.1
86
9.1
14.1
13.2
42.4
7.0
49.8
1.0
45.5
Outliers
removed
106
21.7
18.6
19.5
32.8
5.6
30.0
-0.3
-10.1
85
9.1
13.7
13.1
29.8
6.3
45.8
0.4
44.5
Concentration 2
All
data
113
43.0
62.6
64.1
57.9
9.0
14.3
-1.4
-0.5
92
38.8
36.9
38.0
61.6
10.4
28.1
0.0
-2.2
Outliers
removed
110
64.4
63.5
64.3
44.0
7.2
11.3
-0.4
-0.2
90
38.8
36.8
38.0
53.3
9.4
25.5
-0.2
-2.2
Concentration 3
All
data
112
64.7
64.0
65.0
73.0
10.8
16.8
-1.2
0.5
91
88.9
82.5
81.9
74.3
11.2
13.6
-0.2
-7.9
Outliers
removed
109
64.7
65.1
65.0
58.2
8.4
13.0
0.2
0.5
88
88.9
83.0
82.1
47.6
9.2
11.1
0.5
-7.7
Concentration 4
All
data
112
126.2
120.8
121.3
146.0
16.6
13.7
0.1
-3.9
90
128.3
125.6
125.7
89.9
13.9
11.1
-0.4
-2,0
Outliers
removed
108
126.2
121.6
121.6
83.8
11.8
9.7
0.3
-3.6
86
128.3
125.7
125.7
59.0
10.6
8.4
-0.4
-2.0
Concentration 5
All
data
112
204.0
190.4
197.0
271.3
38.9
20.4
-1.9
-3.4
91
192.2
169.6
169.3
116.3
18.0
10.6
0.0
-11.9
Outliers
removed
109
204.0
194.7
197.5
212.0
29.6
15.2
-0.8
-3.2
89
192.2
169.7
169.3
78.8
15.9
9.4
0.0
-11.9
*A statistic indicating the lack of symmetry in a distribution. For a normal distribution this value is near zero.
Median - True value „ lnn
True value x 10°
-------
Tables 9 and 10 indicated serious problems with Audit 1078 results.
The percent differences for concentrations 1 and 5 were not consistent with
previous values (Table 9). A comparison in Table 10 of the true value and
the median also indicate a discrepancy. In past surveys, the median has
been an excellent indicator of the true value. Skewness values indicate
that the results are normally distributed.
A discrepancy of this magnitude indicates a problem existing with
either the samples or the EPA analysis. An investigation revealed that
for concentrations 1 and 5, corroborative analyses provided by a contractor
yielded results much closer to the survey median. Analysis from both
groups showed an unusually large number of outlier samples. Reanalysis
by EPA after the audit was completed also yielded values nearer the study
median but still quite variable.
Because of the sample variability, it was difficult to determine
whether the EPA analyses were incorrect or a series of bad samples were
analyzed. To prevent further occurrences of this nature, more stringent
acceptance controls are being instituted. Audit results will be reported
for informational purposes only and will not be used in future accuracy
or precision calculations.
Each laboratory data set was plotted against its corresponding EPA
data set, and the slope and intercept of the linear regression line were
determined. For Audit 0478, the mean of 113 slopes was 1.015 with a stan-
dard deviation of 0.234; the mean intercept was 3.90 with a standard devi-
ation of 18.00. Audit 1078 had a mean slope of 1.138 with a standard
23
-------
deviation of 0.132; the mean intercept was -5.52 with a standard deviation
of 10.87. Figure 6 shows a graphical plot of these results.
0.
o
_i
oo
1.10 4
1.05
1.00
0.95
0.90
0.85
INTERCEPT
(-5.52)
ID
O
ID
f--
r~~ co
I s
co
r»-
o
6
.5
-3
.2
.1
CJ>
D-
UJ
O
Figure 6.
i— O
STUDY NUMBER
Plot of S02 slope and intercept.
24
-------
Analytical Method Summary
A check was made to determine whether sets excluded from the summary
were related to the method of analysis (i.e., was one method responsible for
most of the data -outliers, Table 11).
TABLE 11. SUMMARY OF S02 ANALYTICAL METHODS USED BY OUTLIER LABS
Method Total number Number identified Percent of total
using method as outliers as outliers
S02 audit - 0478
Pararosani 1 i ne-manual
Pararosani 1 i ne-automated
Other
S02 audit - 1078
Pararosani 1 i ne-manual
Pararosani 1 i ne-automated
Other
93
29
3
79
21
0
8
2
2
7
1
0
9
7
67
9
5
0
As shown, no one method contributed to the unusually large portion of outlier
data.
To determine whether a particular analytical method produced biased
results, Table 12 was developed. This table contains the mean and standard
deviation of each concentration for each procedure used. No pattern was
established in any of the audits.
Recheck Program
Laboratories reporting at least 3 results greater than +20 percent of the
EPA values were sent a second set of samples. This procedure would distin-
guish between labs having chronic problems and those who just had a "bad
day." For audit 0478, 25 laboratories received a second set of samples. Of
25
-------
ro
TABLE 12. MEAN AND STANDARD DEVIATION OF S02 RESULTS BY ANALYTICAL METHOD (yg/m3)
Cone. 1
Mean Std. dev.
S02 audit - 0478
Manual 20.8 25.6
Automated 17.7 7.0
Other 10.5 0.0
True value 21.7
S02 audit - 1078
Manual 13.9 7.5
Automated 12.4 6.8
True value 9.1
Cone. 2 Cone. 3
Mean Std. dev. Mean Std. dev.
62.2 9.0 64.4
63.8 9.1 64.2
60.3 0.0 21.0
64.4 64.
36.6 10.9 83.3
38.2 8.5 79.5
38.8 88.
9.5
11.6
0.0
7
9.5
16.2
9
Cone. 4
Mean Std. dev.
119.5 18.0
124.1 10.6
141.5 0.0
126.2
125.2 14.4
126.8 12.1
128.3
Cone. 5
Mean Std. dev.
191.5 34.8
185.8 50.2
225.3 0.0
204.0
170.3 16.7
167.2 22.0
192.2
-------
the 11 groups returning data, 5 had corrected their problems to the extent
that at least 3 of their samples fell within +20 percent of the EPA values.
Twenty-one laboratories received recheck samples for audit 1078.
Eleven of these groups returned data, 7 of which had corrected their problems
so that at least 3 samples fell within +20 percent of the EPA values.
Only one agency received recheck samples for both audits. Thus, it
appears that a small percentage of the agencies performing SO,, analysis still
have basic analytical problems. It is also evident that many of the problems
uncovered during the initial analysis were corrected before the recheck
samples were received.
Summary
The S0? audit starting dates covered in this report were April and
October, 1978. The number of participants varied from 100 to 125. Foreign,
EPA, State, local, and private laboratories submitted data.
Three analytical methods were used. The majority of the laboratories
in both audits used the manual pararosaniline method. Aporoximately 22
percent of the laboratories used the automated pararosaniline procedure.
The overall results of audit 0478 showed no bias of any practical signi-
ficance between reported and EPA values. The statistical distribution
appeared to be normal. Large discrepancies in audit 1078 were noted, and
were apparently due to analytical errors or questionable sample materials.
The average slopes (reported vs. EPA) for each audit ranged from 1.105
to 1.138. Intercepts varied from 3.90 to -5.52 yg/m .
AMBIENT NITROGEN DIOXIDE
Participant Characteristics
Nitrogen dioxide audit, number 0678, began in June 1978. Out of 122
-------
sample sets requested by participants, 89 sets of data were returned for a
response rate of 73 percent. Audit number 1278 began in December 1978. Out
of 95 sample sets requested, 77 sets of data were returned for a response
rate of 81 percent.
Table 13 indicates the monitoring agency distribution.
TABLE 13. N02 AGENCY DISTRIBUTION
N02 audit - 0678
Agencies requesting samples
Agencies returning data
N02 audit - 1278
Agencies requesting samples
Agencies returning data
Foreign
1
1
1
0
EPA
4
1
1
0
State
45
37
34
32
Local
55
42
47
38
Private
17
8
12
7
Total
122
89
95
77
Analytical methods were grouped into six broad categories. Table 14
lists the procedures used and their corresponding number of respondents.
TABLE 14. N02 ANALYTICAL METHODS
Method
Saltzman-manual
Saltzman-automated
Sodium arsenite-manual
Sodium arsenite-automated
TGS - ANSA-manual
Other
Agencies
Audit 0678
4
2
59
19
2
3
using method
Audit 0677
4
1
52
17
2
1
28
-------
A laboratory reporting usage of the manual sodium arsenite procedure as the
method of choice may have used that procedure with various modifications.
Thus, in Table 14, 59 laboratories employed a procedure approximating the
manual sodium arsenite method during audit 0678.
Acceptable Ranges
Two performance ranges were used as one means available for judging
performance. The Sample and Target ranges for N02 are listed in Table 15
and apply to sample values in increasing concentrations.
TABLE 15. N02 SAMPLE AND TARGET RANGES(%)
"••••'•-'' -----
Sample range
Target range
Cone. 1
±10
±20
Cone. 2
±10
±20
Cone. 3
±10
±20
Cone. 4
±10
±20
Cone. 5
±10
±20
Neither the Sample nor the Target range was determined using the
method described under Statistical Approach. Both ranges were arbitrarily
set at the values listed in Table 15. As more audits are conducted and more
data becomes available, the ranges will be refined.
Using the above criteria, a tabulation was made of the agencies obtaining
results within the ranges. For audit 0678, a total of 68 (76%) agencies
reported all 5 results within the Target ranges, while 5 (6%} laboratories
reported all results outside these ranges. Corresponding figures for audit
1278 showed 60 (79%) and 4 (5%), respectively. Figure 7 shows a running plot
of those values.
29
-------
80
UJ
CD
w noh-d
in the Appendix by an asterisk (*).
30
-------
Table 16 is a frequency distribution of the percent difference between
the reported and EPA values for each sample concentration. The differences
were calculated by the following formula:
absolute percent difference =
reported value - EPA value
EPA value
x 100.
The frequency distribution was then constructed and appears in Table 16. It
should be noted that the "All samples" line is not an average of the numbers
appearing above it, but is the distribution resulting from the total
compilation of data.
Table 16 is also very useful for laboratories trying to determine their
performance relative to the other participants. For example, in audit
0678, only 10% of the labs reporting results for concentration 1 had a
percent difference of 0.8 or less, while 50% of the laboratories reported
a percent difference of 4.8 or less for the same concentration. The table
also indicates the average percent difference for all laboratories for all
samples. These values are shown in Figure 8 along with corresponding
numbers from previous studies.
-------
TABLE 16. N02 ABSOLUTE PERCENT DIFFERENCE
N02 audit - 0678
Cone. 1
Cone. 2
Cone. 3
Cone. 4
Cone. 5
All samples
N02 audit - 1278
Cone. 1
Cone. 2
Cone. 3
Cone. 4
Cone. 5
All samples
No. Min.
82 0.0
84 0.0
84 0.0
83 0.3
82 0.2
415 0.0
72 0.0
72 0.0
69 0.2
72 0.0
72 0.2
357 0.0
10%
0.8
0.8
' 0.8
1.6
0.9
1.1
0.8
0.3
0.4
0.3
0.6
0.6
30% 50% 70% 90%
3.2 4.8 8.7 15.9
4.2 5.5 8.0 15.6
3.8 5.8 7.5 13.8
3.5 4.6 7.0 9.9
2.0 3.8 5.1 8.7
3.3 5.0 7.5 13.5
3.4 6.8 8.0 13.3
2.1 3.6 5.7 10.6
1.4 3.7 5.6 11.8
1.4 2.9 3.7 7.6
1.3 2.7 3.9 10.4
1.9 3.4 5.7 11.7
Max. Mean
98.4 8.5
55.5 7.5
54.2 7.4
34.0 6.0
43.7 4.9
98.4 6.9
37.9 7.8
19.3 4.9
24.2 5.2
19.9 3.8
51.5 4.5
51.5 5.2
UJ
o
LU
a:
LLJ
u_
u_
1 — 1
a
UJ
ID
O
oo
CO
30-,
20 •
10 •
0
^
<£>
O
^-*~—- ~~
— -^
r-.
CVJ
~^-^
\
r- i>. 60 do ~"
VD C\J ID OJ
O I— O r—
STUDY NUMBER
Figure 8. Plot of N02 absolute percent differences.
32
-------
All data received (except omitted data) were grouped according to con-
centration. The results for each sample are listed in the Appendix in
increasing concentration. The summary statistics which appear in Table 17
and at the top of each listing in the Appendix do not include outliers.
Table 17 lists summary statistics. The "Outliers removed" columns
contain data on which outlier criterion was applied. All data that met the
second criterion stated in Section 3 were removed. It should be noted that
the variation in the number of samples from concentration to concentration is
due to laboratory accidents and damaged samples. With the exception of one
sample (Table 17), all accuracy figures were less than ±5 percent. All
concentrations were normally distributed. The overall accuracy of audit
1278 was superior to audit 0678.
Individual data sets were plotted against corresponding EPA values, and
the slope and intercept of their linear regressions were determined. For
audit 0678, the mean of 84 slopes was 0.988 with a standard deviation of
0.094; the mean intercept was -0.007 with a standard deviation of 0.025.
Audit 1278 had a mean slope of 0.994 with a standard deviation of 0.063;
the mean intercept was -0.001 with a standard deviation of 0.030. Figure 9
illustrates a tally of these results.
33
-------
TABLE 17. N02 SURVEY STATISTICS (yg/ml)
GO
Concentration 1
All Outliers
data removed
N02 audit -
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
N02 audit -
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
0678
82
0.13
0.13
0.13
0.15
0.02
14.10
3.55
2.38
1278
72
0.26
0.27
0.27
0.14
0.03
9.74
1.01
2.55
81
0.13
0.13
0.13
0.08
0.01
9.80
0.82
2.38
70
0.26
0.27
0.27
0.11
0.02
8.30
0.41
2.65
Concentration 2
All Outliers
data removed
84
0.24
0.25
0.25
0.18
0.02
8.90
1.58
5.04
72
0.39
0.39
0.39
0.13
0.02
6.35
0.09
0.52
82
0.24
0.25
0.25
0.10
0.02
5.71
0.15
5.04
71
0.39
0.39
0.39
0.12
0.02
6.03
-0.15
0.52
Concentration 3
All Outliers
data removed
84
0.24
0.25
0.25
0.16
0.02
8.61
. 1.96
4.58
69
0.52
0.52
0.52
0.22
0.04
7.38
-0.73
0.78
83
0.24
0.25
0.25
0.09
0.02
6.99
0.39
4.58
66
0.52
0.52
0.52
0.15
0.03
5.71
0.48
0.97
Concentration 4
All Outliers
data removed
83
0.37
0.39
0.39
0.18
0.02
6.28
1.28
4.02
72
0.70
0.71
0.70
0.26
0.04
5.67
0.72
0.00
80
0.37
0.39
0.39
0.08
0.02
4.43
-0.27
4.02
69
0.70
0.70
0.70
0.18
0.03
4.47
0.82
0.00
Concentration 5
All Outliers
data removed
82
0.69
0.71
0.71
0.43
0.05
7.03
-3.34
3.05
72
0.93
0.94
0.94
0.74
0.08
8.40
-2.75
0.91
80
0.69
0.71
0.71
0.19
0.03
4.22
0.52
3.19
70
0.93
0.94
0.94
0.27
0.05
4.79
0.04
0.91
*A statistic indicating lack of symmetry in a distribution. For a normal distribution this value is near zero.
Median - True value
True value
x 100
-------
1.15,
1.10.
1.05
oo
UJ
o 1.00
oo
0.95
0.90
0.85
vo
O
Intercept
IO
r^.
CM
<£>
O
evj
00
r--.
vo
O
cvj
STUDY NUMBER
. 0.04
0.03
• 0.02 5
O.OT §
01
- 0.00 -1
-0.01
Figure 9.
Plot of N02 slooe and interceot.
Analytical Method Summary
A check was made to determine whether any relationship existed between
35
-------
the laboratories whose data were omitted from the summaries and the analy-
tical method employed (i.e.4; was any one method responsible for most of the
data outliers). The following table (Table 18) resulted.
As can be seen, no particular analytical method was responsible for an
unusually large portion of the outlier data.
TABLE 18. SUMMARY OF N02 ANALYTICAL METHODS USED BY OUTLIER LABS
Method
N02 audit - 0678
Sal tzman-automated
Sodium Arsenite-manual
Sodium Arsenite-automated
Other
N02 audit - 1278
Sal tzman-automated
Sodium Arsenite-manual
TGS-ANSA-manual
Total number
using method
2
59
19
3
1
52
2
Number identified
as outliers
1
2
1
1
1
2
1
Percent of total
as outliers
50
3
5
33
100
4
50
To determine whether a particular analytical method produced biased
results, Table 19 was developed. This table contains the mean and standard
deviation of each sample concentration for each method used. As can be
seen, both major methods estimated the true concentration well and both
were equally precise.
Recheck Program
Laboratories reporting 3 or more results greater than +20 percent of the
36
-------
.co
TABLE 19. MEAN AND STANDARD DEVIATION OF N02 RESULTS BY ANALYTICAL METHOD (yg/ml)
Method Concentration 1 Concentration 2 Concentration 3 Concentration 4 Concentration 5
Mean Std. dev. Mean Std. dev. Mean Std. dev. Mean Std. dev. Mean Std. dev.
N02 audit - 0678
Saltztnan-man.
Saltzman-auto.
Sodium arsenite-man.
Sodium arsenite-auto.
TGS - ANSA-man.
Other
True value
N02 audit - 1278
Saltzman-man.
Sodium arsenite-man.
Sodium arsenite-auto.
TGS - ANSA-man.
Other
True value
0.13
0.13
0.13
0.13
0.13
0.11
0.
0.26
0.27
0.28
0.27
0.32
0.
0.00
_
0.02
0.01
0.02
0.02
13
0.02
0.03
0.03
-
-
26
0.25
0.25
0.25
0.24
0.24
0.23
0
0.38
0.39
0.39
0.38
0.43
0
0.01
M
0.02
0.01
0.03
0.02
.24
0.03
0.02
0.03
-
-
.39
0.25
0.25
0.26
0.25
0.25
0.23
0
0.51
0.52
0.52
0.52
0.55
0
0.01
_
0.02
0.02
0.03
0.02
.24
0.02
0.04
0.05
-
-
.52
0.38
0.40
0.39
0.38
0.39
0.37
0
0.72
0.70
0.70
0.68
0.74
0
0.01
_
0.03
0.02
0.04
0.02
.37
0.06
0.03
0.05
-
-
.70
0.68
0.73
0.72
0.68
0.69
0.68
0
0.97
0.93
0.94
0.92
1.04
0
0.01
-
0.03
0.08
, 0.08
.69
0.05
0.08
0.06
-
-
.93
-------
EPA values was sent a second set of samples. For audit 0678, 7 laboratories
received a second set. Of the 7 groups returning data, 3 had corrected their
problem to the extent that at least 3 samples fell within +20 percent of EPA
values.
Four laboratories received recheck samples for audit 1278. One of these
groups returned data, however, their values remained unacceptable.
Summary
Nitrogen dioxide audit dates were June and December, 1978. The number
of participants varied from 95 to 122. Foreign, EPA, State, local, and
private laboratories submitted their results.
Six analytical methods were employed. In both audits, the majority of
laboratories used the manual sodium arsenite procedure. Approximately 20
percent of the laboratories used the automated sodium arsenite procedure.
Overall results revealed no bias of any practical significance between
the reported and EPA values. The data from each audit appeared to be normally
distributed.
The average slopes (reported vs. EPA) for each audit ranged from 0.98 to
0.99 and intercepts varied from 0.001 to -0.007 yg/ml.
AMBIENT CARBON MONOXIDE
Partici pant Characteri stics
Carbon monoxide audit, number 0378, began in March, 1978. Out of 145
agencies requesting samples, 117 returned data for a response rate of 81
percent. Three hundred twenty-seven instruments were tested. Audit 0978
began in September, 1978. Out of 149 agencies requesting samples, 122
returned data for a response rate of 82 percent. A total of 318 CO monitors
were audited. Table 20 indicates the monitoring agency distribution.
-------
Methods used to analyze the samples were grouped into three categories.
Table 21 lists the analytical methods used and the number of instruments using
the method.
TABLE 20. CO AGENCY DISTRIBUTION
CO Audit - 0378
Agencies requesting samples
Agencies returning data
CO audit - 0978
Agencies requesting samples
Agencies returning data
Foreign EPA State Local
3 8 53 77
3 6 44 63
2 8 59 75
2 4 49 65
Private Total
4 145
1 117
5 149
2 122
TABLE
21 . CO ANALYTICAL METHODS
Method
NDIR
FID
Other
Analyzers using method
Audit 0378
280
42
5
Audit 0978
268
40
4
Most instruments classified as "Other" used an electrochemical
method of detection.
Agencies generally employ a liberal interpretation of the analytical
method used. However, in the case of CO analyses, the methods are clearly
defined. Unknowns possibly effecting results and which are considered part
of the analytical system are the purity of zero air, and the accuracy of
calibration standards.
39
-------
Acceptable Ranges
As described in a previous section, two performance ranges were used as
one means of judging performance. The Sample and Target ranges for the
audits are listed in Table 22 and apply to concentrations in ascending order.
TABLE 22. CO SAMPLE AND TARGET RANGES (%)
CO audit - 0378
Sample range
Target range
CO audit - 0978
Sample range
Target range
Concentration 1
± 7.0
±10
±13.1
±13.1
Concentration 2
± 2.5
±10
± 3.4
±10
Concentration 3
± 1.2
±10
± 1.4
±10
Sample ranges were not determined by the method described under Statis-
tical Approach. Because filling of the cylinders was done so precisely,
standard deviations of the verification analyses were small. Thus, using the
procedure described in the Statistical Approach section resulted in un-
reasonably small Sample ranges for all concentrations. Rather than use
unrealistic values, the Sample ranges were set at a QAD determined value
±0.5 ppm.
Target ranges were determined from previous CO audit results.
Earlier studies indicated that the average percent difference between re-
ported results and EPA determined values for all concentrations was +10
percent. This value has been used for all Target ranges with the exception
of concentration 1. In that particular case, the Sample range was aoplied
due to its larger value.
40
-------
A tabulation was made of the instruments reporting results within the
ranges. For audit 0378, a total of 222 instruments (68%) reported all values
within the Target ranges, while 5 instruments (2%) reported all results
outside the Target ranges. Corresponding values for audit 0978 showed 166
(52%) and 11 (3%), respectively. Figure 10 shows a graphical plot of those
values.
60
50
£ 40
30
20
10
0
vo i£>
r-» r^~
«a- o
o r—
ALL 3 SAMPLES IN
TARGET RANGES
ALL 3 SAMPLES OUTSIDE
TARGET RANGES
cr>
o
oo
r-»
ro
o
oo
i —
01
o
STUDY NUMBER
Figure 10. Plot of CO results (3 samoles within and outside Target ranges)
41
-------
Data Summary
Any instrument not reporting at least one value within the Target ranges
was considered an outlier and was not included in the data summaries. The
remainder of the discussion will deal with the reported results minus the
values identified as outliers. These results are eliminated from further
data summaries. The values identified as outliers are indicated in the
Appendix by an asterisk (*).
Table 23 is a frequency distribution of the percent difference between
the reported and EPA values for each sample concentration. The differences
were calculated by the following formula:
absolute percent difference =
reported value - EPA value
x 100.
EPA value
The frequency distribution was then constructed and appears in Table 23. It
should be noted that the "All Samples" line is not an average of the numbers
appearing above it, but is the distribution resulting when all data is
examined together regardless of concentration.
Table 23 assists laboratories in determining their relative performance.
For example, in audit 0378 only 10% of the instruments reporting results for
concentration 1 had a percent difference of 1.4% or less, while 50% of the
instruments reported a difference of 4.9% or less for the same concentration.
The table also indicates the average percent difference for all instruments
for all samples. These values are illustrated in Figure 11 along with
corresponding values from previous audits.
42
-------
TABLE 23. ABSOLUTE PERCENT DIFFERENCE
Concentration
No.
Hin.
10%
30%
50%
70%
90%
Max.
Mean
CO audit - 0378
All
1
2
3
samples
319
321
317
957
0.1
0.1
0.1
0.1
1.4
0.4
0.2
0.4
2.1
1.1
0.8
1.3
4.9
1.8
1.6
2.3
9.1
3.8
3.1
4.9
16.1
7.3
6.4
11.0
555
19
88
555
.9
.9
.7
.9
9.4
3.2
3.1
5.2
CO audit - 0978
All
1
2
3
samples
301
301
301
903
0.3
0.0
0.1
0.0
2.1
0.4
0.1
0.4
4.7
1.9
1.2
2.1
9.3
3.0
2.0
3.8
16.2
5.8
3.2
7.1
30.1
9.3
6.2
16.8
161
198
54
198
.8
.1
.7
.1
14.1
5.1
2.9
7.4
UJ
o
40 -
30
s 20 -
10 •
o
CO
CQ
VD
r»-
«d-
o
10
r--
o
ro
CTi
O O
STUDY NUMBER
00
!-».
CO
o
CO
I--
cn
o
Figure 11. Plot of CO absolute percent differences,
43
-------
The results for each sample value are listed in the Appendix, in
increasing concentration. The summary statistics which appear in Table 24
and at the top of each listing in the Appendix do not include outliers.
Table 24 lists summary statistics based on the reported data. The
"Outliers removed" column contains data on which the outlier criterion was
applied. All data that met the second criterion in Section 3 were removed.
It should be noted that the variation in the number of samples from concen-
tration to concentration is due to laboratory accidents and damaged samples.
An examination of Table 24 reveals no bias or skewness problems in any
of the audits. The greatest inaccuracy was -6.5 percent. The near zero
value of the skewness indicator shows that the distribution was normal.
The EPA determined true value and the median of the study results
agree well. This is indicated by accuracy values. The study population is
normally distributed as indicated by the skewness.
Each instrument data set was plotted against its corresponding EPA
data set, and the slope and intercept from the linear regression were
determined. For audit 0378, the mean of 322 slopes was 0.973 with a
standard deviation of 0.175; the mean intercept was 0.545 with a standard
deviation of 3.773. Audit 0978 had a mean slope of 0.978 with a standard
deviation of 0.064; the mean intercept was 0.261 with a standard deviation
of 1.277. A total of 307 instrument results were used to determine these
values. Figure 12 represents a tally of these results.
-------
TABLE 24. CO SURVEY STATISTICS
Concentration 1
CO audit - 0378
Number
True value
Mean
Medi an
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
CO audit - 0978
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
All
data
319
7.2
7.1
7.0
42.9
2.4
33.4
15.1
-2.1
301
3.8
3.6
3.6
8.5
0.8
22.1
2.9
-6.0
Outliers
removed
318
7.2
6.9
7.0
6.1
0.7
10.7
-0.3
-2.1
297
3.8
3.6
3.6
4.5
0.6
16.8
0.5
-6.5
Concentration 2
All
data
321
20.2
20.3
20.3
7.2
0.9
4.6
-0.6
0.4
301
14.6
14.8
14.7
32.1
1.9
12.7
11.8
1.0
Outliers
removed
318
20.2
20.4
20.3
6.0
0.9
4.3
-0.2
0.5
300
14.6
14.7
14.7
7.7
0.9
6.0
0.3
1.0
Concentration 3
All
data
317
42.1
42.2
42.3
42.6
3.3
7.8
-8.6
0.5
301
36.4
36.9
36.9
25.5
1.8
4.8
-4.9
1.2
Outliers
removed
315
42.1
42.4
42.3
13.6
1.6
3.7
-0.5
0.6
300
36.4
37.0
36.9
9.4
1.3
3.5
0.5
1.2
*A statistic indicating a lack of symmetry in a distribution.
distribution this value is near zero.
For a normal
median - true value
true value
x 100
45
-------
UJ
Q.
O
1.10
1.05'
1.00-
0.95-
0.90-
0.85
SLOPE
VO VD
t*» Is-
"*• o
O i—
n
O
01
O
00
r^
co
O
00
r>»
cr>
O
STUDY NUMBER
Figure 12. Plot of CO slone and intercept.
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Q_
O_
Q.
UI
O
Analytical Method Summary
A check was made to determine whether any one method was responsible
for most of the data outliers and resulted in Table 25. As can be seen,
both of the predominately used procedures contributed only 5 percent or
less of the outlier instruments.
To determine whether a particular analytical method produced biased
results, Table 26 was developed. This table contains the mean and standard
deviation of each sample concentration for each method employed.
46
-------
TABLE 25. SUMMARY OF CO ANALYTICAL METHODS USED BY OUTLIER INSTRUMENTS
Method
Total no.
using method
No. identified
as outliers
Percent of total
as outliers
CO audit - 0387
NDIR
Flame ionization
Other
CO audit - 0978
NDIR
Flame ionization
Other
280
42
5
268
40
4
5
0
0
9
2
1
2
0
0
3
5
25
TABLE 26. MEAN AND STANDARD DEVIATION OF CO RESULTS BY ANALYTICAL METHOD (PPM)
Method
Concentration 1
Mean Std. dev.
CO audit - 0378
NDIR
Flame ionization
Other
True value
CO audit - 0978
NDIR
Flame ionization
Other
True value
7.0
7.3
7.5
7.2
3.7
3.4
6.1
3.8
0.8
6.3
0.9
0.7
0.6
3.4
Concentration 2
Mean
20.4
19.7
20.7
14.7
15.1
16.0
Std. dev.
0.8
1.6
0.8
20.2
0.8
4.8
2.6
14.6
Concentration 3
Mean
41.8
41.8
42.2
37.0
36.5
37.2
Std. dev.
5.1
6.1
3.0
42.1
1.2
3.9
2.4
36.4
47
-------
The data Indicate that the NDIR and FID methods were equally accurate;
however, the NDIR was more precise. The "Other" category was biased high,
however, only 4 or 5 instruments submitted data.
Summary
The CO audit commencement dates covered in this report were March and
September, 1978. Participants varied from 145 to 149; up approximately 3.0%
from the previous year. Number of instruments checked varied from 318 to
327; also up approximately 3.0%. Foreign, ERA,, State, local and private
laboratories submitted data.
Three analytical methods were utilized. The majority of the instruments
were NDIRs (approximately 85 percent).
Overall results showed no bias between the reported and EPA values,
and results appeared to be normally distributed. Of the three procedures
used, the NDIR and FID techniques seemed to yield equally accurate
results. However, NDIR appeared to be slightly more precise. Average
slopes (reported vs. EPA) for each audit varied from 0.973 to 0.978
with intercepts from 0.261 to 0.545.
HI-VOL SULFATE
Participant Characteristics
Sulfate audit number 0278 began in February, 1978. Out of 73 sample
sets requested by participants, 49 sets of data were returned for a response
rate of 67 percent. Audit 0878 began in August, 1978. Out of 84 sample
sets requested, 58 sets of data were returned for a response rate of 69
percent.
Table 27 compiles the monitoring agency type distribution.
48
-------
Methods used to analyze the samples were grouped into 9 broad categories.
Results were received from laboratories using 5 methods and are listed in
Table 28.
TABLE 27. S0| AGENCY DISTRIBUTION
Foreign
SOj audit - 0278
Agencies requesting samples
Agencies returning data
SOJ audit - 0878
Agencies requesting samples
Agencies returning data
2
1
2
2
EPA
1
1
3
2
State
30
20
32
26
Local
23
16
26
17
Private
17
11
21
11
Total
73
49
84
58
TABLE 28. SOj ANALYTICAL METHODS
Method
Methyl thymol blue - automated
Barium chloride - manual
Barium chloride - automated
Sulfa-ver - manual
Other
Agencies
Audit 0278
16
21
1
8
3
using method
Audit 0878
18
25
3
9
3
It should be noted that some agencies tend to define the analytical methods
used in very general terms. A laboratory reporting usage of the automated
methyl thymol blue procedure may have used that procedure with various
modifications. Thus, Table 28 should be interpreted as 16 laboratories in
audit 0278 having used procedures approximating the automated methyl thymol
blue method.
49
-------
Acceptable Ranges
As described in a previous section, two performance ranges were used
as one means of judging performance. The Sample and Target ranges are
listed in Table 29 and apply to sample concentrations in ascending order.
TABLE 29. S0° SAMPLE AND TARGET RANGES (%)
Sample range
Target range
Cone. 1
± 5
±15
Cone. 2
± 5
±15
Cone. 3
± 5
±15
Cone. 4
± 5
±15
Cone. 5
± 5
±15
Cone. 6
± 5
±15
Neither the Sample nor Target ranges were determined using Statistical
Approach methods. Both were arbitrarily set at the values listed in Table
29. As more audits are conducted and more data become available, the
ranges will be refined.
Using those criteria, a tabulation was made of the agencies reporting
results within the ranges. For audit 0278, 12 (24%) agencies reported 5
or 6 results within the Target ranges, while 5 (10%) laboratories reported
all results outside these ranges. Corresponding figures for audit 0878
showed 19 (33%) and 3 (5%), respectively. Figure 13 illustrates a graphi-
cal plot of those values.
Data Summary
Using the Target ranges as one means of eliminating nonrepresentative
data, it was decided that any laboratories not reporting at least one value
within the Target ranges were considered outliers. Labs performing in-
adequately were excluded from the summaries. Five laboratories met the
criterion for outlier rejection in 0278 and 3 in 0878.
50
-------
60,
50
40
CO
UJ
o 30
UJ
CD
fe 20
10
5 OR 6 SAMPLES
IN TARGET RANGE
ALL SAMPLES OUTSIDE
RGET RANGES
rv rv
tv. rv
CM CO
O O
00
rv
CM
o
CO
rv
CO
o
STUDY NUMBER
Figure 13. Plot of S0| results (5 or 6 samples within Target ranaes and
6 samples ciutside Target ranges).
Table 30 is a frequency distribution of the percent difference between
reported and EPA values for each concentration. The differences were
calculated using the following formula:
absolute percent difference =
reported value - EPA value
EPA value
x 100.
The frequency distribution was then constructed and appears in Table 30.
It should be noted that the "All Samples" line is not an average of the
numbers appearing above it, but is the distribution resulting when all data
is examined together regardless of concentration.
51
-------
Table 30 is helpful to laboratories trying to determine their relative
performance. For example, only 10 percent of the labs in audit 0278 reporting
results for concentration 3 had a percent difference of 3.3 percent or
less, while 50 percent of the laboratories reported a percent difference of
12.2 percent or less. In addition, the table also indicates the average
percent difference for all laboratories for all samples (except concentration
1). These values are tabulated in Figure 14 along with corresponding
values from previous audits.
Concentration T was a blank for audit 0278 and was excluded in the
"All samples" distribution. Because small concentration differences result
in large percent differences, it was felt that those particular values would
unduly distort the study results.
TABLE 30. SOj ABSOLUTE PERCENT DIFFERENCE
Concentration No. Min. 10% 30% 50% 70% 90% Max. Mean
SOJ audit - 0278
1
2
3
4
5
6
All samples 215 0.0 0.8 4.4 8.9 15.6 48.2 521.7 20.9
SO audit - 0878
BLANK
41
43
43
44
44
0.0
0.0
0.0
0.0
0.0
3.9
3.3
0.2
0.3
0.3
11.1
7.4
2.1
1.9
3.2
19.4
12.2
5.0
5.4
6.5
44.4
27.8
10.7
8.5
9.9
88.9
55.6
19.7
21.7
15.6
521.7
191.1
95.0
94.9
94.8
47.3
28.2
11.0
9.8
9.9
1
2
3
4
5
6
All samples
44
52
54
51
53
52
306
0.0
0.0
0.0
0.0
0.0
0.3
0.0
4.6
3.0
1.4
1.7
0.6
0.9
1.5
20.0
6.7
4.6
4.0
3.0
3.6
4.6
36.2
17.7
8.1
6.0
4.6
7.3
4.1
53.8
30.0
11.7
11.1
11.9
10.2
18.6
167.7
65.0
25.0
25.6
21.2
16.7
41.5
423.8
205.0
46.0
43.6
41.8
37.0
423.8
72.0
28.7
10.9
11.2
9.0
9.0
22.1
52
-------
60,
40-
*s 30j
UJ
tr
o 20J
QO
10-
c\j
o
oo
o
oo
r~-
c\j
o
oo
r--.
oo
o
STUDY NUMBER
Figure 14. Plot of SO^ absolute percent differences,
Results for each sample value are listed in the Appendix in order of
increasing concentrations. The summary statistics in Table 31 and at the
top of each listing in the Appendix do not include outliers.
Table 31 lists summary statistics based on reported data, and the
"Outliers removed" column contains data on which an outlier te-.t w,r. ,t|>|tli<4
-------
An examination of Table 31 shows that all the data were normally
distributed. Inaccuracy at some of the concentrations was larger than
would be desired. However, most of the bias occurred at low concentrations.
Generally, there tended to be a positive bias in the lower concentrations
and a negative bias in the higher concentrations.
Individual laboratory data sets were plotted against their corres-
ponding EPA data sets, and the slopes and intercepts of the linear repression
line were determined. For audit 0278, the mean of 43 slopes was 1.031 with
a standard deviation of 0.152; the mean intercept was -0.383 with a standard
deviation of 1.353. Audit 0878 had a mean slope of 1.050 with a standard
deviation of 0.151; the mean intercept was -0.324 with a standard deviation
of 1.510. Figure 15 shows a plot of these results.
(1.41)
1.15
1.10
1.05 -I
00
UJ
§5 1-00
oo
0.95
0.90
0.85
0.50
0.00
CD
o
cc
Lul
-0.50 g
-1.00
CM
o
oo
o
oo
r~-
CM
o
oo
r->.
oo
o
STUDY NUMBERS
Figure 15. Plot of S07 slope and intercept.
54
-------
TABLE 31. SOj SUMMARY STATISTICS (yg/m3)
01
Concentration 1
All Outliers
data removed
SOj audit -
Number
True value
Mean
Medi an
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
SOJ audit -
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
0278
29
0.00
1.26
0.51
6.59
1.55
122.91
1.90
-
0878
44
1.30
2.08
1.68
6.12
1.34
64.48
1.87
29.23
28
0.00
1.07
0.50
4.60
1.18
110.52
1.63
-
42
1.30
1.87
1.64
4.41
0.94
50.47
1.38
26.15
Concentration 2
All Outliers
data removed
41
1.80
2.26
1.98
11.01
1.68
74.16
3.66
10.00
52
3.00
3.47
3.15
8.15
1.31
37.60
1.88
5.00
40
1.80
2.04
1.98
4.43
0.89
43.51
0.52
10.00
50
3.00
3.28
3.12
4.25
0.90
27.28
0.19
4.00
Concentration 3
All Outliers
data removed
43
2.70
2.91
2.90
7.68
1.23
42.25
1.46
7.41
54
11.10
10.98
10.94
8.37
1.62
14.73
-0.36
-1.40
42
2.70
2.80
2.85
5.74
0.97
34.71
0.25
5.74
53
11.10
11.08
10.95
6.96
1.48
13.33
0.05
-1.35
Concentration 4
All Outliers
data removed
43
9.60
9.22
9.59
11.12
2.00
21.72
-2.86
-0.10
51
13.50
12.71
12.90
8.76
1.98
15.60
-0.69
-4.44
41
9.60
9.60
9.60
4.25
0.94
9.76
-0.01
-0.00
51
13.50
12.71
12.90
8.76
1.98
15.60
-0.69
-4.44
Concentration 5
All Outliers
data removed
44
12.00
11.56
11.88
14.06
2.16
18.70
-2.92
-0.96
53
26.40
25.52
25.65
17.10
3.25
12.73
-0.44
-2.84
43
12.00
11.82
11.89
6.52
1.37
11.55
-0.42
-0.92
52
26.40
25.72
25.67
13.56
2.95
11.46
-0.01
-2.75
Concentration 6
All Outliers
data removed
44
14.40
13.92
14.02
17.52
2.58
18.55
-2.87
-2.60
52
39.60
38.62
38.36
25.29
21.58
12.03
-0.27
-3.12
43
14.40
14.23
14.05
9.34
1.61
11.34
0.03
-2.43
51
39.60
38.88
38.40
22.77
4.27
10.97
0.11
-3.03
*A statistic indicating the lack of symmetry in a distribution. For a normal distribution this value is near zero.
median - true value
true value
x 100
-------
Analytical Method Summary
A check was made to determine whether any one method was responsible
for most of the data outliers (Table 32).
TABLE 32. SUMMARY OF SOjj ANALYTICAL METHODS USED BY OUTLIER LABS
Method
S0| audit - 0278
Methyl thymol blue -
automated
Barium chloride - manual
Sulfa-ver - manual
S0| audit - 0878
Barium chloride - manual
Total no.
using method
15
18
7
22
No. identified
as outliers
1
3
1
4
% of total
as outliers
7
17
14
18
While no specific method was responsible for a large percentage of the
outlier data, the manual barium chloride method was responsible for outlier
data in both of the audits. Table 33 contains the means and standard
deviations of each sample concentration for each method used.
In general, the automated methythymol blue procedure was superior in
accuracy and precision. There was little difference in the accuracies of
both barium chloride methods and the manual Sulfa-ver method. Automated
barium chloride was more precise than manual barium chloride. None of the
methods indicated a large bias.
Recheck Program
Starting with audit 0278, a recheck program was begun. Any laboratory
reporting at least 3 results greater than ±15 nercent of the EPA values was
56
-------
TABLE
33. MEAN AND STANDARD
DEVIATION OF SOj RESULTS BY ANALYTICAL METHOD (yg/nr)
Method
SOJ audit - 0278
MTB - auto.
Bad - man.
Bad - auto.
Sulfa-ver - man.
Other
True value
S0| audit - 0878
MTB - auto.
Bad - man.
Bad - auto.
Sulfa-ver - man.
Other
True value
Cone
Mean
0.55
1.34
0.40
0.48
0.20
0.
1.46
2.04
1.20
1.71
1.61
1.
. 1
Std.
dev.
1.14
1.76
0
0.84
0.26
00
0.76
2.04
0.70
1.24
0.23
30
Cone
Mean
2.53
2.28
2.60
1.35
1.26
1.
3.26
3.65
3.85
3.14
2.89
3.
. 2
Std.
dev.
2.44
1.16
0
1.00
0.94
80
0.66
2.00
1.49
0.69
0.14
00
Cone.
Mean
3.02
3.05
4.20
2.19
1.89
2.70
10.90
10.85
12.33
11.19
10.48
11.10
3
Std.
dev.
1.39
1.08
0
1.40
1.50
0.86
2.08
2.18
1.54
0.45
Cone.
Mean
8.98
9.92
8.50
9.43
6.18
9.60
12.76
12.56
12.02
13.23
2.89
13.50
4
Std.
dev.
1.96
1.24
0
0.70
4.94
1.19
2.39
3.89
1.66
0.14
Cone. 5
Mean
11.85
11.93
11.20
11.66
7.79
12
24.14
25.68
29.21
26.54
25.37
26
Std.
dev.
1.12
1.74
0
1.17
6.23
.00
3.16
3.15
3.16
3.10
0.10
.40
Cone.
Mean
14.21
14.43
13.40
14.08
9.19
14.
36.95
39.14
41.21
39.43
38.76
39
, 6
Std.
dev.
0.87
2.23
0
1.33
7.34
40
4.14
5.17
4.53
4.78
1.98
.60
-------
sent a second set of samples. For audit 0278, 24 laboratories received a
second set of samples. Of the 16 groups returning data, 4 had corrected
their problems to the extent that at least 3 of their samples fell within
±15 percent of the EPA values.
Thirty-one laboratories received recheck samples for audit 0878. Nine-
teen of these groups returned data, of which 12 had corrected their problems
to the point where at least 3 samples fell within ±15 percent of the EPA
values.
From our compiled results, we concluded that a large percentage of the
laboratories were having S0| analytical problems. The recheck samples
indicated that the difficulties were real and that the poor performance was
not due to chance.
Summary
The SO^ audit start dates covered in this report are Febuary and
August, 1978. The number of participants ranged from 73 to 84. Foreign,
EPA, State, local, and private laboratories submitted data.
Five analytical methods were used with the majority of laboratories
employing either the automated methylthymol blue, manual barium chloride, or
manual Sulfa-ver methods.
Overall results showed no bias between the reported and EPA values and
statistical distributions appeared to be normal. The automated methylthymol
blue method exhibited the greatest accuracy and precision. The barium
chloride and Sulfa-ver methods were approximately equal in accuracy.
Average slopes (reported vs. EPA) for each audit ranged from 1.031 to 1.050
with intercept values from -0.383 to -0.324.
58
-------
HI-VOL NITRATE
Participant Characteristics
Nitrate audit number 0278 began in February, 1978. Out of 49 labora-
tories returning sulfate data, all submitted nitrate results; two groups
analyzed only nitrate. Audit 0878 began in August, 1978. Out of 58 sulfate
results received, 47 returned nitrate data. Table 34 indicates the moni-
toring agency distribution.
Methods used to analyze the samples were grouoed into 5 broad categories,
Table 35 lists the analytical methods used and the number of respondents for
each method.
TABLE 34.
AGENCY DISTRIBUTION
Foreign
EPA State Local
Private Total
W^ audit - 0278
Agencies
requesting samples
Agencies
returning data
Ml audit - 0878
2
2
23
23
12
12
13
13
51
51
Agencies
requesting samnles
Agencies
returning data
2
2
2
2
21
21
9
9
13
13
47
47
59
-------
TABLE 35. NO" ANALYTICAL METHODS
Method
Cadmium reduction -
Cadmium reduction -
Hydrazine reduction
Hydrazine reduction
Other
manual
automated
- manual
- automated
Aaencies
Audit 0278
4
23
1
6
17
using method
Audit 0787
4
22
1
7
13
It should be noted that some agencies tend to define the analytical methods
used in very general terms. A laboratory reporting the use of the automated
cadmium reduction method may have used that method with various modifica-
tions. Thus, Table 35 should be interpreted as 23 laboratories in audit
0278 having used procedures approximating the automated cadmium reduction
method.
Acceptable Ranges
As described in a previous section, two performance ranges were used as
one means of judging performance. The Sample and Target ranges for N03 are
listed in Table 36 and apply to concentrations in ascending order.
TABLE 36. N0~ SAMPLE AND TARGET RANGES (%)
Sample range
Target range
Cone. 1
± 5
±15
Cone. 2
± 5
±15
Cone. 3
± 5
±15
Cone. 4
± 5
±15
Cone. 5
± 5
±15
Cone. 6
± 5
±15
60
-------
Both ranges were arbitrarily set at the values listed in Table 36. As
more audits are conducted and more data become available, the ranges will be
refined.
For audit 0278, a total of 12 (24%) agencies reported 5 or 6 results
within the Target ranges, while 2 (4%) laboratories reported all results
outside these ranges. The corresponding figures for audit 0878 showed 24
(51%) and 7 (15%), respectively. Figure 16 shows a graph of the above
values.
Data Summary
Any laboratory not reporting at least one value within the Target
ranges would be considered an outlier and its data was not included in the
summaries. Two laboratories met this criterion for outlier rejections in
audit 9278, and 7 in 0878.
60i
50
.„ 40-1
30
20
10
0
UJ
I—I
o
5 OR 6 SAMPLES
IN TARGET RANGES
ALL SAMPLES OUTSIDE
TARGET RANGES
—r—
10
r>. r-. oo
r-- r-- r-.
CM co CM
co
|~r
oo
STUDY NUMBER
Figure 16. Plot of N(£ results (5 or 6 samples within and 6 samples
outside Target ranges).
61
-------
Table 37 is a frequency distribution of the percent difference between
the reported and EPA values for each sample concentration. The differences
were calculated by the following formula:
absolute percent difference =
reported value - EPA value
EPA value
x 100.
The frequency distribution was then constructed and appears below. It
should be noted that the "All samples" line is not an average of the numbers
appearing above, but is the data examined together regardless of the
concentration.
TABLE 37. NO^ ABSOLUTE PERCENT DIFFERENCE
Concentration
No.
Win.
10%
30%
50%
70%
90%
Max .
Mean
NO, audit - 0278
1
2
3
4
5
6
All samples
BLANK
49
48
48
49
48
242
0.0
0.0
0.0
0.0
0.1
0.0
17.8
0.7
0.2
0.5
0.4
0.8
33.3
4.2
1.9
1.7
1.9
2.9
60.0
13.2
5.7
3.6
4.2
8.0
86.7
18.8
9.7
7.4
8.3
20.0
151
39
16
20
22
77
.1
.6
.4
.4
.5
.8
1766
650
43
.72
89
1766
.7
.0
.0
.7
.9
.7
141.0
43.1
7.9
8.8
11.8
42.8
N03 audit - 0878
1
2
3
4
5
6
All samples
40
40
38
40
39
39
236
0.0
0.0
0.0
0.0
0.2
0.0
0.0
1.4
0.0
0.0
0.8
0.6
1.0
0.6
8.3
4.4
2.1
2.5
1.9
2.1
2.7
18.1
8.3
4.2
6.7
3.8
6.0
6.1
38.9
14.4
6.0
7.2
5.8
9.6
11.1
62.5
32.8'
15.0
16.0
16.8
18.4
31.9
166.7
490.6
43.8
36.7
38.5
108.8
490.6
30.7
24.1
7.1
7.6
6.7
11.0
14.6
62
-------
Table 37 enables laboratories to determine their relative performance.
For example, only 10% of the labs in audit 0278 reported results for
concentration 3 that had a percent difference of 0.7 or less, while 50% of
the laboratories reported a percent difference of 13.2 or less for the same
concentration. The table also indicates the average percent difference for
all laboratories for total samples. These values are illustrated in
Figure 17 along with corresponding values from previous audits.
60-1
50-
£40'
30
o:
LU
t-
o
20
10
0
CM
O
co
O
CO
r--
oo
o
STUDY NUMBER
Figure 17. Plot of NOZ absolute percent differences,
63
-------
Concentration 1 was a blank for audit 0278 and was excluded from the
"All samples" distribution. Because small concentration differences result
in large percent differences, it was felt that the numbers would unduly
distort the study results. Attention should be given to the mean for 0278.
The 42.8 percent value is inflated due to the low concentration of sample 2.
If sample 2 is omitted, the mean is approximately 14 percent.
Results for each sample value are listed in the Appendix in increasing
concentration. The summary statistics which appear in Table 38 and at the
top of each listing in the Appendix do not include outliers. The "Outliers
removed" column contains data on which an outlier test was applied. All
data that met the second criterion described in Section 3 were removed. It
should be noted that the variation in the number of samples from concen-
tration to concentration is due to laboratory accidents, damaged samples,
and inclusion of duplicate samples.
Table 38 indicates that the results were normally distributed (skewness
near zero). Lower value samples appeared to cause accuracy problems.
However, in general the accuracy was within 5 percent.
Each set of laboratory data was plotted against its corresponding EPA
data set, and the slope and intercept of the linear regression line were
determined. For audit 0278 the mean of 49 slopes was 1.047 with a standard
deviation of 0.216; the mean intercept was -0.185 with a standard deviation
of 2.409. Audit 0878 had a mean slope of 1.013 with a standard deviation of
0.159; the mean intercept was 0.008 with a standard deviation of 0.565.
Figure 18 shows a plot of these results.
64
-------
TABLE 38. NO^ SUMMARY STATISTICS (pg/irT)
Concentration 1
All
data
Outliers
removed
Concentration 2
All
data
Outliers
removed
Concentration 3
All
data
Outliers
removed
Concentration 4
All
data
Outliers
removed
Concentration 5
All
data
Outliers
removed
Concentration 6
All
data
Outliers
removed
171
10~ audit - 0278
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
NOl audit - 0878
28
0.00
0.57
0.24
8.99
1.66
290.76
4.62
27
0.00
0.26
0.21
0.92
0.21
81.20
1.07
49
0.45
1.04
0.72
8.30
1.51
144.73
4.19
60.00
47
0.45
0.74
0.69
1.79
0.11
44.80
1.04
53.33
48
1.
1.
1.
44
90
50
10.49
1.82
95.82
4.08
4.17
46
1
1
1
3
.44
,54
.50
,34
0.45
29.01
1.80
4.17
48
10.50
10.33
10.40
7.68
1.22
11.77
-0.63
-0.95
46
10.50
10.35
10.40
4.45
0.94
9.05
-0.32
-0.95
49
12.00
11.57
11.81
12.32
1.89
16.31
-1.97
-1.58
47
12.00
11.85
11.83
7.76
1.27
10.72
-0.25
-1.42
48
14.40
13.17
14.21
14.74
3.27
24.84
-2.51
-1.35
45
14.40
13.91
14.25
8.23
1.56
11.19
-1.82
-1.04
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
40
0.72
0.88
0.80
1.65
0.30
34.18
1.32
11.11
38
0.72
0.83
0.80
1.00
0.21
25.44
-0.08
11.11
40
1.80
2.09
1.86
9.53
1.42
67.74
5.44
3.61
39
1.80
1.87
1.85
1.70
0.31
16.56
0.54
2.78
38
4.80
4.66
4.76
2.87
0.56
11.97
-1.69
-0.83
36
4.80
4.76
4.78
1.80
0.35
7.41
0.03
-0.42
40
6.00
5.80
5.86
3.16
0.62
10.69
-1.12
-2.25
38
6.00
5.90
5.90
2.26
0.47
7.95
-0.24
-1.67
39
9.60
9.30
9.45
5.31
0.95
10.23
-1.24
-1.56
38
9.60
9.39
9.49
3.97
0.78
8.30
-0.45
-1.09
39
11.40
11.61
11.16
14.62
2.56
22.00
3.30
-2.11
37
11.40
11.09
11.13
4.92
1.01
9.11
0.49
-2.37
*A statistic indicating the lack of symmetry in a distribution. For a normal distribution this value is near zero.
median - true value
true value
x 100
-------
2.25,
2.00.
1.75-
1.50.
oo
UJ
o 1.25.
1.00-
0.75
INTERCEPT
[•>.
**•
c
r-. r--
OJ 00
o o
STUDY NUMBER
r
oo
0.50
CO
o.oo
D-
UJ
O
UJ
b -0.50 -
-1.00
00
r-.
oo
o
Figure 18. Plot of NO" slope and intercept.
Analytical Method Summary
A check was made to determine whether any relationship existed between
the laboratories that submitted data that was omitted from the summaries and
the analytical method employed (i.e., was any one method responsible for
most of the data outliers). Table 39 resulted from this check.
No particular method was determined responsible for most of the outlier
values. (See Table 39). It is also obvious that, as yet, no one method is
the "method of choice".
66
-------
TABLE 39. SUMMARY OF NO" ANALYTICAL METHODS USED BY OUTLIER LABS
Method T9tai no. No. identified % of total
_ using method as outliers as outliers
NO" audit - 0278
Cadmium reduction -
manual 4 00
Cadmium reduction -
automated 23 0 o
Hydrazine reduction -
manual 1 1 -|00
Hydrazine reduction -
automated 6 00
Other 17 1 6
NOg audit - 0878
Cadmium reduction -
manual 4 00
Cadnium reduction -
automated 22 3 14
Hydrazine reduction -
manual 1 1 100
Hydrazine reduction -
automated 7 1 14
Other 13 2 15
Table 40 contains the means and standard deviations of each sample
concentration for each method used and is useful in determining whether
a particular method yielded biased values.
Recheck Program
Starting with audit 0278, a recheck program was begun. A laboratory
reporting at least 3 results greater than ±15 percent of the EPA values
was sent a second set of samples. For audit 0278, 18 laboratories received
a second set. Of the 5 groups returning data, 2 had corrected their problems
to the extent that at least 3 samples fell within ±15 percent of the EPA
67
-------
values. Fourteen laboratories received recheck samples for audit 0878. Nine
of these groups returned data, of which 5 had corrected their problems.
TABLE 40. MEAN AND STANDARD DEVIATION OF NOj RESULTS BY ANALYTICAL GROUP
Method Cone. 1 Cone. 2 Cone. 3 Cone. 4 Cone. 5 Cone. 6
xsxsxsxsxsxs
NO^ audit - 0278
Cad. red. -
manual 2.25 4.50 3.60 4.12 3.98 4.56 10.601.42 13.161.72 15.200.87
Cad. red. -
automated 0.14 0.18 0.72 0.39 1.82 1.76 10.421.52 11.481.41 12.863.71
Hyd. red. -
automated 0.37 0.32 0.77 0.36 1.88 0.93 10.45 0.59 11.13 1.66 13.68 1.40
Other 0.10 0.15 0.78 0.27 1.48 0.25 10.09 0.87 11.46 2.48 12.89 3.47
True value 0.00 0.45 1-44 10.50 12.00 14.40
HQ~3 audit - 0878
Cad. red. -
manual 0.72- 0.27 1.84 0.58 4.72 0.41 5.381.06 8.291.66 13.936.66
Cad. red. -
automated 0.86 0.30 2.33 2.04 4.71 0.34 5.84 0.38 9.43 0.67 11.34 1.94
Hyd. red. -
automated 0.87 0.21 1.84 0.22 4.08 1.00 5.44 0.77 8.90 1.09 11.26 1.65
Other 0.97 0.36 1.92 0.17 4.90 0.40 6.09 0.60 9.69 0.75 11.42 1.02
True value 0.72 1.80 4.80 6.00 9.60 11.40
Summary
The N03 audit start dates covered in this report were February and
August, 1978. The number of participants submitting data ranged from 47
to 51. Foreign, EPA, State, local, and private laboratories returned
results.
Four analytical methods were used (plus a category, "Other"). A l,m)o
oercentage of the samples were analyzed by methods in the "Other" category.
68
-------
The overall results showed, with the exception of very low concentrations,
little bias between the reported and EPA values. The statistical distribu-
tion of the results appeared to be normal. The automated cadmium reduction
method showed the best accuracy, while the automated hydrazine reduction
procedure had the best precision.
Average slopes (reported vs. EPA) for each audit ranged from 1.01 to
1.05 with intercepts from -0.185 to 0.008.
HI-VOL LEAD
Participant Character!' sti cs
Lead audit number 0178 began in January, 1978. Out of 85 samole sets
requested by participants, 69 sets of data were returned for a response rate
of 81 percent. Audit 0678 began in June, 1978. Out of 88 samole sets
requested by participants, 67 sets of data were returned for a response rate
of 76 percent. Table 41 indicates the monitoring agency distribution.
TABLE 41. PB AGENCY DISTRIBUTION
Foreign
Pb audit - 0178
Agencies requesting samples
Agencies returning data
Pb audit - 0678
Agencies requesting samples
Agencies returning data
0
0
0
0
EPA
5
2
6
5
State
42
37
42
35
Local
30
25
30
23
Private
8
5
10
4
Total
85
69
88
67
Methods used to analyze the samples were grouped into 4 broad categories,
Results were received from laboratories using 2 methods listed in Table 42.
69
-------
TABLE 42. PB ANALYTICAL METHODS
Method Agencies using method
Audit 0178Audit 0678
Atomic absorption 69 68
Other 0 1
It should be noted that some agencies tend to define the analytical
methods used in very general terms. A laboratory reporting usage of the
atomic absorption procedure as the method of choice may have used that
procedure combined with modifications. Thus, Table 42 should be interpreted
as 69 laboratories having used procedures during audit 0178 approximating
the atomic absorption method.
A tabulation of the methods used to extract material from the filter
is in Table 43.
TABLE 43. PB EXTRACTION PROCEDURES
Method Agencies using method
Audit 0178Audit 0678
Hot acid extraction 62 58
Cold acid extraction 3 3
Ultrasonication 1 3
Other 3 5
Acceptable Ranges
Sample and Target ranges are listed in Table 44 and apply to samole
concentrations in ascending order. Both ranges were arbitrarily set at the
values listed in Table 44. As more audits are conducted and more data
become available, the ranges will be refined.
70
-------
TABLE 44. PB SAMPLE AND TARGET RANGES (%)
Sample range
Target range
Cone. 1
± 5
±10
Cone. 2
± 5
±10
Cone. 3
± 5
±10
Cone. 4
± 5
±10
Cone. 5
± 5
±10
Coric. 6
± 5
±10
For audit 0178, 29 agencies (51%) reported 5 or 6 samples within the
Target ranges, while 7 laboratories (10%) reported results outside the
Target ranges. The corresponding figures for audit 0678 showed 29 (43%) and
4 (6%), respectively.
Data Summary
Laboratories not reporting at least one value within the Target ranges
were considered outliers and excluded from the summaries. Seven labora-
tories were rejected during audit 0178 and 4 during audit 0678; they were
thus eliminated from further data summaries. The values identified as
outliers are indicated in the Appendix by an asterisk (*).
Table 45 is a frequency distribution of the percent difference between
the reported and EPA values for each sample concentration. The differences
were calculated by the following formula:
absolute percent difference =
reported value - EPA value
J IL L ~—~"—~J~~~ " .•.iiia • I • !• • -•
EPA value
x 100.
The frequency distribution was then constructed and appears below. It
should be noted that the "All Samples" line is not an average of the number
appearing above it, but is the distribution resulting when all data is
examined together regardless of concentration.
71
-------
TABLE 45. PB ABSOLUTE PERCENT DIFFERENCE
Concentration
Pb audit - 0178
1
2
3
4
5
6
All samples
Pb audit - 0678
1
2
3
4
5
6
All samples
No.
Min.
10%
30%
50%
70%
90%
Max .
Mean
BLANK
62
61
61
62
61
307
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.2
0.6
0.8
0.5
3.
2.
2.
1.
1.
2.
3
4
4
7
6
4
6.7
3.9
4.6
3.9
4.7
4.8
15.0
8.3
7.1
7.1
7.9
8.3
35.0
13.2
13.8
11.8
12.3
18.3
230
78
79
89
81
230
.0
.9
.1
.2
.5
.0
17.6
8.0
7.6
8.6
7.9
10.0
BLANK
62
62
62
63
63
312
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.7
0.6
0.3
0.0
0.4
2.
3.
1.
2.
2.
2.
4
3
8
2
8
8
4.8
5.3
4.2
4.2
5.0
4.8
11.9
8.0
6.1
8.4
7.0
8.0
31.0
15.3
12.5
24.3
15.9
20.9
969
68
79
81
79
969
.0
.0
.2
.0
.8
.0
27.7
8.7
7.1
10.6
8.4
12.5
Table 45 is useful to laboratories in determining their relative
performance. For example, only 10% of the labs reporting results for
concentration 4 during audit 0178 had a percent difference of 1.2 or less,
while 50 % of the laboratories reported a percent difference of 4.6 or
less for the same concentration. The table also indicates the average
percent difference for all laboratories for total samples. These values are
tabulated in Figure 19.
Concentration 1 was a blank and was excluded in the "All samples"
distribution. Because small concentration differences result in large per-
cent differences, it was felt that the numbers would unduly distort the
study results.
72
-------
LU
O
UJ
50
40
30
20
UJ
CO
-------
TABLE 46. Pb SUMMARY STATISTICS (yg/m3)
Concentration 1
All Outliers
data removed
Pb audit -
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
Pb audit -
Number
True value
Mean
Median
Range
Std. dev.
Coef. var.
Skewness*
Accuracy
0178
67
0.00
0.30
0.00
1.20
2.20
733.30
3.12
-
0678
69
0.00
0.20
0.00
9.70
1.20
600.00
3.18
-
66
0.00
0.05
0.00
0.40
0.10
200.00
1.60
-
68
0.00
0.04
0.00
1.20
0.30
850.00
2.10
-
Concentration 2
All
data
62
0.6
0.6
0.6
1.8
0.2
35.6
3.5
0.0
66
0.4
0.5
0.4
4.4
0.5
106.2
6.5
0.0
Outliers
removed
60
0.6
0.6
0.6
0.8
0.1
20.1
-0.6
0.0
65
0.4
0.4
0.4
1.7
0.2
44.2
4.9
0.0
Concentration 3
All
data
61
2.0
2.0
2.0
2.0
0.3
15.8
-3.2
0.0
66
1.5
1.5
1.5
1.8
0.2
16.1
-0.6
1.0
Out! i ers
removed
59
2.0
2.0
2.0
0.9
0.2
8.0
-0.4
0.0
64
1.5
1.5
1.5
1.1
0.2
12.0
-0.1
1.0
Concentration 4
All
data
61
6.6
6.4
6.4
7.3
0.9
13.7
-2.7
-2.7
66
5.0
4.8
4.9
6.1
0.8
16.2
-2.0
-1.8
Outliers
removed
60
6.6
6.5
6.4
3.6
0.6
9.1
0.7
-2.5
63
5.0
4.9
4.9
2.6
0.4
9.0
-0.4
-1.8
Concentration 5
All
data
62
10.2
9.7
10.0
13.4
1.8
18.6
-2.9
-1.6
67
9.9
9.7
9.9
13.1
2.1
21.8
-1.3
0.0
Outliers
removed
60
10.2
10.0
10.0
8.8
1.0
10.0
0.2
-1.4
65
9.9
9.9
9.9
11.0
1.6
16.5
0.0
0.0
Concentration 6
All
data
61
12.6
11.9
12.1
14.8
1.8
14.9
-2.7
-3.8
67
12.0
11.3
11.6
15.6
2.1
18.2
-1.8
-3.0
Outliers
removed
58
12.6
12.1
12.1
5.3
0.8
6.7
-0.6
-3.7
64
12.0
11.5
11.6
5.6
1.2
10.1
-0.8
-3.0
*A statistic indicating the lack of symmetry in a distribution. For a normal distribution this value is near zero.
median - true value
true value
x 100
-------
With the exception of concentration 1, the EPA determined "True value",
and the mean and median of the study results agree well. This is indicated
in the "accuracy" column. The greatest difference was approximately 3.7
percent. The relatively poor agreement between the EPA and reported values
for concentration 1 is an indication of the minimum detectable limits of the
analytical methods used.
Each data set was plotted against its corresponding EPA set, and the
slope and intercept were determined. For audit 0178, the mean of 62 slopes
was 1.120 with a standard deviation of 0.612; the mean intercept was -0.052
with a standard deviation of 0.409. The corresponding values for audit 0678
were a slope of 1.068 with a standard deviation of 0.534, and an intercept
of 0.014 with a standard deviation of 0.468.
Analytical Method Summary
A check was made to determine whether any relationship existed between
the five laboratories that submitted data that were omitted from the
summaries and the analytical method employed (i.e., was any one method
responsible for most of the data outliers). Table 47 resulted from this
check.
As can be seen in Table 47, with the exception of one laboratory, all
of the outlier labs used the atomic absorption method.
Recheck Program
Starting with audit 0178, a recheck program was begun. Any laboratory
reporting at least 3 results greater than ±10 percent of the EPA values
was sent a second set of samples. For audit 0178, 24 laboratories received
a second set of samples. Of the 5 groups returning data, only 1 had corrected
it problems to the extent that at least 3 samples fell within ±10 percent
of the EPA values.
75
-------
TABLE 47. SUMMARY OF PB ANALYTICAL METHODS USED BY OUTLIER LABS
Method Total number
using method
Pb audit
Atomi c
Other
Pb audit
Atomi c
Other
- 0178
absorption 69
0
- 0678
absorption 68
1
Number identified
as outliers
7
0
4
0
% of total
as oul tiers
10
0
6
0
Twenty-five laboratories received recheck samples for audit 0678.
Eleven of these groups returned data, of which 5 had corrected their problems
to the extent that at least three samples fell within ±10 percent of the EPA
values.
Summary
The Pb audit survey 0178 began in January, 1978. Eighty-five labora-
tories requested samples, of which 69 returned results for a response rate
of 81 percent. Audit 0678 began in June, 1978. Eighty-eight laboratories
requested samples, of which 67 returned results. EPA, State, local, and
private laboratories submitted data. Two analytical methods were used.
Ninety-nine percent of the laboratories used the atomic absorption procedure.
One laboratory used the "Other" technique.
Overall agreement between reported and EPA values was good. No bias
was evident, and the results v/ere normally distributed. The average percent
difference between the EPA results and reported results was 10.0 percent
76
-------
for audit 0178, and 12.5 percent for audit 0678. This figure does not
include the blank data.
HI-VOL FLOW RATE
Participant Characteristics
Hi-vol flow rate audit number 0578 began in May, 1978. Out of 221
agencies requesting to participate, 162 responded with data, for a response
rate of 73 percent. A total of 1,241 Hi-Vol units were tested.
Table 48 indicates the monitoring agency distribution. Methods used
to measure the flow rate were grouped into 3 categories: rotameters,
pressure transducers, and "other". Table 49 lists the measurement methods
and the number of units employing the method.
TABLE 48. HI-VOL FLOW RATE AGENCY DISTRIBUTION
Foreign
Hi-vol audit 0578
Agencies requesting ReF 4
Agencies returning data 4
EPA State Local Private
9 69 127 12
5 54 94 5
Total
221
162
TABLE 49. HI-VOL FLOW
Method
Rotameter
Pressure transducer
fH-hai*
RATE MEASUREMENT METHODS
Units usinq method
Audit 0578
496
530
215
77
-------
Table 49 indicates that, compared to previous years, a trend is
developing away from the rotameter to other more precise and accurate
measurement methods.
Acceptable Ranges
Sample and Target ranges were calculated as +5 and +9 percent,
respectively.
Using this criterion, a tabulation was made of the number of samplers
reporting results within the ranges. A total of 744 samplers (60%) reported
4 or 5 readings within the Target ranges, while 144 units (12%) reported
4 or 5 readings outside the Target ranges. Fifty agencies accounted for 111
units reporting all values outside the Target ranges.
Data Summary
Table 50 is a frequency distribution of the percent difference between
the reported and EPA values for each measurement pair. The differences were
calculated using the following formula:
percent difference = ^ported^value^EPA value .
An iterative routine eliminated outliers while constructing the
distribution. During each pass of the data, a check was performed to
determine which reported values met the second criterion mentioned in
Section 3. These values were removed, and the procedure repeated until no
outliers were identified. This is the distribution that 1s titled "Outliers
removed" in Table 50. A total of .8 passes were required to remove all
outliers; 185 values were omitted.
78
-------
TABLE 50. HI-VOL FLOW RATE PERCENT DIFFERENCES
No. Min. 10% 30% 50%
Hi-vol audit - 0578
All values 5983 -513 -12 -4 -1
Outliers removed 5798 -24 -10 -4 -1
70% 90% Max. Mean
2 6 100 -18
2 6 21 -1
Removal of the outliers had little effect on the main body of the Table.
Table 50 is useful when evaluating the overall audits. Excluding
outliers, of all the observations reported for audit 0578, 80 percent were
within +10 percent of the EPA calculated value. The overall average
».
difference was -1 percent.
A histogram was constructed of the total values (Figure 19) and
reveals a slight negative bias (values less than EPA). The distributions
appear normal, with slight negative skewing. However, considering the
diversity of the measurement sources, the results appear well behaved and
showed excellent precision.
All reported value pairs for each audit were summarized using linear
regression equation:
y = mx + b (5)
where y = reported value
x = EPA value
The resulting equation for audit 0578 was:
y = 0.953 x +2.609 (6)
Because of the large intercept, the equation appears to indicate a larger
bias than actually exists. Since equation 6 is valid only over the range
79
-------
HI-
^.«.
r <: . i
OV.1
tt>.i
f C,
iv. f - ?i
5t.5 -'
iJ.* -!
5C.Z -M
W.1 -M
tl-.V ^71
*...>• -•«:
37.F
3*.i
3T.TT
"75TT ~ =T»
21. V -V
"T1«7E~ - 1)(
15. t -'.n
TZT5
)l XTT5f* '
»x»xrxxxxxxxxxxxxxxxxxxxrxxxrxxxxxxx*x*xxxxxxxxxxxxxxxxxr
. »XXXXIXX»XXX»XXXXXXXXXXXXXXXXXXXXXXXtXXXXX»XXXXX1'XXX1fXXX»XX''
iTrx"JfHf XX X'O'XXTXXX'Vi »'X »X'» X » X yjr » i( y »» )r »X » X XXX X »X X X '
?X»X*XXXXXXXXXXXX»X*XII)'X*)'XX«X »X*
ca -iZV? - :x xxxxxxxi
-!xtxxxxxxxxxxxxi
- rx
•TV
. I
•TV
-fc.* •!«!
•=75.5 -re
EACH » REP^t StN TS I^WALIHS
Figure 20 . Hi-vol flow rate histooram - 1978
-------
for which the test was conducted (30-60 ft3/min), the influence of the
intercept was not as great as indicated.
Measurement Method Summary
To determine whether a measurement method produced biased results and
to gain an indication of its accuracy and precision, the numbers in Table 50
were separated by method, yielding Table 51.
TABLE 51. SUMMARY OF HI-VOL FLOW RATE MEASUREMENT METHODS
Hi-vol audit - 0578
Rotameter
Pressure
Transducer
No.
2269
2515
Min.
-30
-20
10%
-14
-8
30%
-6
-3
50%
-2
0
70%
1
2
90%
7
6
Max.
25
19
Mean
-2.6
-0.6
The numbers generated in Table 51 are the result of several iterations
similar to the "Outliers removed" values of Table 50.
It is obvious from Table 51 that the pressure transducer method is more
accurate than the rotameter method. The table also supports the idea that
the apparent bias revealed in Table 50 is largely due to the negative bias
of the rotameter readings.
Forty-nine percent of the units using the rotameter were able to
report 4 or 5 values within the Target ranges, while 65 nercent of the units
employing the pressure transducer reported 4 or 5 values within the Target
ranges. Linear regression equations for each method were derived using
equation 4 as described in the preceding sections. The resulting
equations appear below:
81
-------
Equations Measurement Method
y = 0.945 + 3.413 rotameter (7)
y = 0.960 + 1.907 pressure transducer (8)
Pressure transducer measurement pairs resulted in a slope closer to unity
and an intercept closer to zero than did rotameter pairs.
Summary
The Hi-vol flow rate audit covered in this report started in May, 1978.
The number of participants requesting an audit device was 221; 162 returned
data for a response rate of 73 percent. A total of 1,241 samplers were
checked. Foreign, EPA, State, local and private laboratories submitted
data.
The results from both audits showed a slight bias between EPA and
reported results with slightly skewed distribution patterns. Considering
the number of units checked and the number of participating personnel,
bias and skewing were assumed insignificant.
The slope of the equation representing all values was 0.953 with an
intercept of 2.609.
82
-------
SECTION 6
REFERENCES
1. Chauvenet, William. A Manual of Spherical and Practical Astronomy.
J.B. Lippincott & Co., Philadelphia, Pennsylvania, 1863.
2. Code of Federal Regulations (40 CFR) 50.11, Appendix A, pp. 5.11.
3. Federal Register, Vol. 38, Mo. 110, June 8, 1973, pp. 15175-15176.
4. Bromberg, S., Bennett, B., and Lampe, R. Summary of Audit Performance
Measurement of S02, N02, Sulfate, Nitrate - 1976. EPA 600/4-78-004,
January, 1978.
83
-------
APPENDIX
-------
INTER-LABORATORY STUDY
POLLUTANT - S04
SAMPLE NUMBER - Q
" 53
TRUE-VALUE 26.40
1E*N 25.52
"EDIAN 25.65
DATA IN ASCENDING ORDER
678
UNITS - MICR06RAMS PER CUBIC METER
RANGE
VARIANCE
STD. BEV.
COEF. VAR
17.10
TO.55
3.25
f 2.73
C.l.CUPPfcR> 26.40
C.I.CLOWE R) 24.65
SKEWNESS -.44
ACCURACY -2.84
5.460
11.350
15.36
15.50 •
18.90
19.50
20.40
20.72
21.12
21.48
22.72
•22.80
22.80
23.10
23.25
23.97
24.39
24.48
24,75
25.00
25.20
25.20
25.30
25.35
25.40
25.44
25.44
25.62
25.65
25.65
25.70
25.90
25.93
26.10
26.16
36.28
26.32
26.40
26.49
26,58
26. 70
26.88
27.30
27.30
27.4Q
27.6Q
27.60
27.89
28.00
28.77
29.00
30.60
30.60
31.65
32.00
32.46
100.36§
84
-------
INTER-LABORATORY STUDY
POLLUTANT - S04
SAMPLE NUMBER - 2
R 52
TRUE-VALUE 39.60
HEftN 3&.6Z
MEDIAN 38.36
&ATA IN ASCENOINS ORDER
UNITS - HICR06RAHS PER CUBIC HETER
RANGE
VARIANCE
STD. DEV.
COEF. VAR.
25.29
21.58
4.65
12.03
C.1. 39.88
C.I.CLOWE R> 37.35
SKEWNESS -.27
ACCURACY -3.12
8.540
18.92
21.15
24.96
27.48
31.10
32.50
33.00
34.05
34.08
34.13
34.20
34.68
36.00
36.29
36.30
36.34
36.36
36. 48
36.60
37.20
37.48
37.53
37.68
37.8C
37.97
38.16
38.18
38.33
38.40
38.52
39.00
39.36
39.37
39.72
39.86
39.95
40.30
40.30
40.5o
40.6Q
41.34
41.40
42.QO
42.20
42.50
43.20
43.20
43.62
44.70
45.00
45.00
45.30
47.72
50.25
154.790
85
-------
INTER-LABORATORY STUDY
POLLUTANT - SO^
SAMPLE NUMBER - 4
1 54
TRUE-VALUE 11.10
WEAN 10*98
*EDIAN 10.94
DATA IN ASCENDING ORDER
878
UNITS - KICR06RANS PER CUBIC METER
RANGE
VARIANCE
STD. DEV.
COEF. VAR.
8.37
2.62
1.62
14.73
C.I.(UPPER) 11.41
C.I.CLOWE R> 10.55
SKEWNESS -.36
ACCURACY -1.40
1.B3
2.52
5.55
6.00
7.41
7.95
8.22
8.93
9.30
9.60
9.90
10.00
10.01
10.05
10.05
10.07
10.06
10.11
10.14
10.20
10.35
10.37
10.40
10.44
10.59
10.62
10.80
10.83
10.85
10.94
10.95
11.04
11.10
11.20
11.25
11.27
11.33
11.39
11.5Q
11.60
tt.7Q
11.76
11.79
11.80
11.94
12.QO
12.40
12.48
12.5Q
12.60
12.80
13.00
13.50
13.62
13.88
14.10
14.37
43.70f
86
-------
INTER-LABORATORY STUDY
POLLUTANT - SO*
SMPLE NUMBER - 5
* 51
TRUE-VALUE 13.50
JEAN 12.71
1EDIAN 12.90
DATA IN ASCENDING ORDER
678
2.7Q
4.20
6.46
7.62
8.00
8.23
9.10
9.12
10.04
10.05
10.62
10.71
I
11.60
11.82
11.94
12.00
12.25
12.27
12.30
12.34
12.60
12.60
12.69
12.70
UWITS - HICROGRAHS PER CUBIC METER
RANGE
VARIANCE
STD. DEV*
COEF. VAR
8.76
3.93
1.98
15.60
C.I.CUPPER) 13.26
C.I.CLOVE R) 12.17
SKEtfNESS -.69
ACCURACY -4.44
12.80
12.8(j
12.83
12.89
12.90
12.96
13.01
13.14
13.2Q
13.24
13.27
13.38
13.40
13.50
13.50
13.9Q
13.95
13.98
™«QO
14. Q4
14.10
14.25
14.30
14.50
14.59
14.84
15.80
16.00
16..26
16.3S
53.92
87
-------
INTER-LABORATORY STUDY
POLLUTANT - S04
SAWPLE NUMBER - 7
N
TRUE-VALUE
IEAN
MEDIAN
52
3.00
3.47
3.15
DATA IN ASCENDING ORDER
678
UNITS - M1CROGRAMS PER CUBIC METER
RANGE
VARIANCE
STO. DEV.
COEf* WAR
8.t5
1.70
t.31
37.60
C.I.(UPPER) 3.63
C.I.CLOWE R) 3.12
SKEWNESS 1.88
ACCURACY 5.00
1.00
1.20$
1.50
3.10
2.29
2.30
2.34
2.40
2.46
2.50
2.57
2.63
2.76
2.80
2. hi
2.&3
2.86
2.88
2.90
2.91
3.00
3.00
3.01
3.03
3.10
3.1Z
3.12
3.18
3.20
3.20
3.30
3.38
3.45
3.45
3.53
3.62
3. &0
3.80
3.9o
3.96
4.00
4-00
4.Q9
4.20
4.32
4.50
4.SO
4.95
4.95
5.10
5.25
7.15
9.15
11.68 §
-------
INTER-LABORATORY STUDY
POLLUTANT - SO*
SAPPLE NUMBER - 9
678
TRUE-VALUE
HEDIAN
DATA IN ASCENDING ORDER
44
1.30
2.D8
1.68
UNITS - HICR06RAMS PER CUBIC METER
RANGE
VARIANCE
STB. DEV.
COEF. VAR
6.12
1.80
t.34
64.48
C.I.CUPPER} £.48
C.I.(LOWE R) 1.68
SKEMNESS 1.87
ACCURACY 29.23
.24
.45
.69
.77
.83
.90
.90
.91
.98
1.10
1.19
1.29
1.30
1.32
1.36
1.40
1.43
1.44
1.45
1.50
1.53
1.56
1.60
1.72
1.76
1.80
1.80
1.84
1.90
2.00
2.10
2.4Q
2.4Q
2.40
2.6Q
2.70
2.88
3.10
3.48
3.83
3.90
5.10
6.15
6.81
9.05
89
-------
I MIL s -LA?OR ATOR ¥ STUDY
POLLUTANT - $04
SAMPLE NUMBER - 0
TRUE-VALUE
43
2.70
2.91
2.9G
DATA IN ASCENDING ORDER
276
UNITS - MICROGRAMS PtR CU^IC METFR
RAN&E
VARIANCE
STO. DEV.
COEF. VAR
7.68
1*52
1.23
42.25
C.I.(UPPER) 3.28
C.I.
-------
INTER-LABORATORY STUDY
POLLUTANT - S04
SAMPLE NUMBER - 2
TRUE-VALUE
SEAN
41
1.80
2.26
1.98
DATA IN ASCENDING ORDER
278
UNITS - MICROGRftMS PER CUBIC METER
RANGE
VARIANCE
STD. DEV.
COEF. VAR
11.01
2.81
1.68
74.16
C.I. 2.77
C.I.CLOWE R) 1.75
SKEWNESS 3.66
ACCURACY 10.00
.18
.24
.60
.90
.98
.99
1.20
1.25
1.34
1.40
1.45
1.59
1.60
1.70
1.72
1.79
1.30
1.82
1.87
1.88
1.89
1.98
1.98
2.00
2.01
2.03
2.10
2-10
2.11
2.13
2.i2
2.40
2.52
2.58
2.6Q
2.70
2.84
3.22
3.40
3.70
3.75
4.50
4.61
7.15
10.20
11.19
-------
I NTER -L ABOR ATOR Y STUDY
DOLLUTANT - S 04
SAMPLE NUMBER - 3
TRUfc-VALUE 9.60
*!£Afc 9.22
« E D 1 A N 9.59
DATA IN ASCENDING ORDER
UNITS - MICR06RAWS PER CUBIC tfETER
RANGF
VARIANCE
STB. £>EV.
CGEF. VAR.
11.12
4.00
2.00
21.72
C.I.(UPPER) 9.81
C.1.
-------
INTER-LABORATORY STUDY
POLLUTANT - S 04
SAKPLE NUMBER - 4
M 44
TRUE-VALUE 12. DO
11.56
11. ?8.
UNITS - KICR06RAHS PER CUBIC
RANGE
VARIANCE
STD. DEV.
COEF. VAR
14.06
4.67
2.16
18.70
C.I.CUPPER) 1Z.20
C.I.(LOWE R) 10.9?
SKEWNESS -2.92
ACCURACY -.96
DATA IN ASCENDING ORDER
.61
3.15
8.64
8.85
9.30
9.60
9.75
10.56
to.se
10.98
11.07
11.11
11.18
11.2.0
11.20
11.35
11.51
11.55
11.62
11.70
11.70
11. 72
11.77
11.88
11.89
11.89
11.96
12.00
12.00
12.00
12.04
12.06
12.10
12.14
12,2o
12.30
12.6Q
12.87
12.98
13.08
13.44
13.56
13.80
13.85
14.31
14.60
14.67
15.30
93
-------
INTER -LABOR ATORY STUDY
POLLUTANT - S04
SAMPLE NUMBER - 5
TRUE-VALUE 14.40
1EAN 13.92
HEDIAK 14.02
DATA IN ASCENDING ORDER
278
UNITS - MICRO&RftMS PER CUBIC *tT£R
R AN6F
VARIANCE
STD. DEV.
COEF. VAR
17.52
6.67
2.55
18.55
C.I.CUPPtR) 14.68
C.1•(L OW t R) 13.16
SKEWNESS -2.87
ACCURACY -2.60
.75
8.93
9.60
11.00
12.15
12.35
12.76
12.90
12.94
12.96
12.98
13.23
13.24
13.40
13.45
13.47
13.49
13.50
13.76
13.76
13.8*
13.93
13.94
U.OG
14.OS
14.2C
14.23
14.25
14.35
14.36
14.40
14.40
14.40
14.46
14.76
15.40
15.49
15. 6Q
16.02
16.25
16.95
17.27
17.28
18.00
18.22
18.27
^•r
•
94
-------
INTER-LABORATORY STUDY
POLLUTANT - S04
SAPPLE NUMBER - 6
N
278
TRUE-VALUE
CEAN
•tEOIftN
DATA IN ASCENDING ORDER
29
.00
1.26
.51
UNITS - MICROGRftHS P£R CUBIC
RANGE
VARIANCE
STD. DEV.
COEF . VAR
6.59
2.41
1.55
122.91
C.I.(UPPER) 1.83
C.I.(LOWE R) .70
SKfcWWESS 1.90
ACCURACY .00
.01
.12
.12
.15
.25
.28
.36
.43
.43
.48
.48
.49
.50
.50
.51
.51
.55
.90
1.09
1.25
1.40
1.45
1.500
1.50
2.29
2.3o
2.96
4.11
4.50
4.61
6.60
10.80
95
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 7
878
TRUE-VALUE
1EDIAN
DATA IN ASCENDING ORDER
.72
.88
.80
UNITS - H1CROGRAMS PER CUBIC METER
RANGE
VARfANCE
STD. DEV.
COEF. VAR
t.65
.09
.30
34.18
C.I.(UPPER) .97
C.I.(LOy£ R) .76
SKEUNESS 1.3?
ACCURACY 11.11
.70
.70
.71
.71
.71
.72
.74
.75
.75
.77
.78
>7&
,80
,80
i 8u
,80
,84
,85
,85
,89
.93
.95
1*00
1.02
1.08
1.10
1.12
1-22
1.27
1.300
1.70
1.92
2.60§
96
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 0
I 39
TRUE-VALUE 11.40
REftN 11.61
MEDIAN 11.16
DATA IN ASCENDING ORDER
878
UNITS - HICROGRAHS PER CUBIC METER
RANGE
VARIANCE
STD. OEV.
COEF. VAR.
14.62
6.53
2.56
22.00
C.I.CUPPER) 12.42
C.I.CLOWE R) 10.81
SKEWNESS 3.30
ACCURACY -2.11
2.48
2.55
2.55
2.78
5.t6
5.90
9.18
9.30
9.41
9.90
9.98
9.99
10.04
10.30
10.44
10.62
10.63
10.70
10.70
10.71
10.7ft
11.1fc
11.10
11.13
11.13
11.16
11.16
11.19
It.22
11.40
11.50
11.51
11.55
11.57
11.6Q
11.63
11.74
11.ao
12.10
12.18
12.67
13.07
14.10
18.87
23.80
35.98
97
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 2
•I 39
TRUE-VALUE 9.60
1EAN 9.30
«I*N * 9.45
DATA IN ASCENDING ORDER
878
2.11
2.19
2.22
2.32
3.24
4.10
5.90
7.24
7.BO
7.83
j
UNITS - HICR06RAMS PER CUBIC METER
RANGE
VARIANCE
SID. DEV.
COEF. VAR
5.31
.91
.95
10.23
8.46
E.63
8.67
8.71
8.88
9.04
9.06
9. 09
9.10
9.13
9.30
9.3C
9.36
9.42
9.44
9.45
9.54
9.55
5,62
9.66
C«I. (UPPER) 9.60
C.I.(LOWE R) 9.01
SKEWNESS -1.24
ACCURACY -1.56
9.66
9.66
9.71
9.71
9.78
9,79
9.8Q
9.84
9.97
10.00
10.10
10.10
10.60
10.80
11.21
29.42§
98
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 4
TRUE-VALUE
JEAN
"IE&IAN
40
6.00
5.50
5.86
DATA IN ASCENDING ORDER
878
UNITS - MICROGRAMS PER CUBIC METER
RAN6E
VARIANCE
STD. DEV.
COEf. VAR
3.t6
.35
.62
10.69
C.I.(UPPER) 5.99
C.I.CLOWE R) 5.61
SKEWNESS -1.12
ACCURACY -2.25
5.05
5.40
5.43
5.55
5.5?
5.57
5.58
5.58
5.60
5.60
5.66
5.74
5.76
5.77
5.85
5.85
5.88
5.92
5.95
5.98
5.99
6.00
6.Q6
6.06
6.to
6.12
6.30
6.36
6.38
6.40
6.40
6.43
6.43
6.46
6.61
6.96
17.72
99
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 5
»» 38
TRUE-VALUE 4.80
*E*N 4.66
MEDIAN 4.76
DATA IN ASCENDING ORDER
878
UNITS - HICR06RAMS PER CUP1C HETER
RANGE
VARIANCE
STD. DEV.
COEF. VAR
2.67
.31
.56
11.97
C.I.CUPPER) 4.84
C.I.CLOWE R) 4.49
SKEWNESS -1.69
ACCURACY -.83
4.44
4.50
4.51
4.53
4.53
4.54
4.58
4.60
4.60
4.66
4.66
4.67
4.76
4.75
4.77
4.79
4.8C
4.8o
4.80
4.81
4.81
4.9Q
4.90
4.9Q
4.91
4.92
4.92
5.07
5.20
5.20
5.43
5.52
5.57
100
-------
INTER-LABORATORY STUOV
POLLUTANT - N03
SAMPLE NUMBER - 9
878
TRUE-VALUE
KEAN
KEDIAN
40
1.80
2.09
1.86
DATA IN ASCENDIN& ORDER
.41
.41
.42
.63
1.00
1.09
1.10
1.30
1.50
1.55
i
UNITS - MICROGRAMS PER CUBIC METER
RANGE
VARIANCE
STO. DEV.
COEF. VAR
9.53
2.Q1
1.42
67.7*
1.5&
1.60
1.65
1.65
1.65
1.72
1.73
1.75
1.80
1.80
1.8G
I.SG
1.80
1.82
1.85
.8S
.88
.90
• 90
.91
C.I.(UPPER) 2.53
C.I.CLOWE R) 1.65
SKEVNESS 5.44
ACCURACY 3.61
1.91
1.91
1.97
2-00
2.02
2.02
2.06
2.07
2.07
2.08
2.13
2.16
2.39
2.470
2.64
2.80
10.63
101
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER - 0
a/8
TRUE-VALJE
XEDIAN
DATA IN ASCENDING ORDER
49
.45
1.D4
.72
UNITS - WICROGRAHS PER CUBIC METER
RANGE
VARIANCE
STO. DEV.
COEF.
8.30
2.27
1.51
144.73
C.I.(UPPER) 1-46
C.I.(LOWE ft) .6?
SKEWNESS 4.19
ACCURACY 60.00
.10
.18
.?1
.36
.37
.43
.45
.50
.53
.53
.54
55
.56
,56
,60
.60
,60
,63
,63
,66
,67
,68
,69
,69
.72
,72
,72
,76
.78
,8C
,81
.81
,81
.84
.36
.87
.88
.«0
1-00
1-03
1.Q3
1.08
1.08
I,t3
1-3Z •
1.49
1.52
1.89
4.34 §
7.80
8.40
102
-------
INTER-LABORATORY STUDY
POLLUTANT - N03
SAMPLE NUMBER -
278
TRUE-VALUE
MEAN
UNITS - HICR06RAMS PER CUBIC HETER
48
1.44
1.90
1.50
DATA IN ASCENDING ORDER
RANGE
VARIANCE
STD. DEV.
COEF.
10.49
3.32
1.82
95.82
C.I.CUPPER) 2.42
C.I.CLOWE R) 1.39
SKEWNESS 4.08
ACCURACY 4.17
.31
.34
.90
.93
.97
1.20
1.23
1.25
1.30
1.34
1.35
1.40
1.40
1.4D
1.40
1.40
1.42
1.43
1.44
1.44
1.45
1.46
1.47
1.50
1.5C
1.50
1.50
1.51
1.53
1.58
1.59
1.63
1.64
1.65
1.68
1.7Q
1.70
1.71
1.77
1.F3
1.86
1.90
1.94
2.01
2.10
3.65
9.75
10.80
103
-------
INTER-LABORATORY STU&Y
POLLUTANT - wo3
SAMPLE NUMBER - 3
N 48
TRUE-VALUE 14.40
JEAh 13.17
1EDIAN 14.21
DATA IN ASCENDING ORDER
276
UNITS - MICR06RAWS PER
METER
RANGE
VARIANCE
STf>. DEV.
COEF. WAR
14.74
10.70
3.27
?4.84
C.I.COPPER) U.09
C.I.(LOWE R) 12.24
SKEWNF&S -2.51
ACCURACY -1.35
1.46
1.52
3.22
6.*0
7.97
9.42
11.16
11.16
11.S2
12.99
13.08
13.13
13.18
13.25
13.4E
13.68
13.72
13.80
13.80
13.82
13.82
1*,0&
14.12
14.16
14.20
14.21
14.25
14 . 2 &
14.30
14.35
14.41
14.44
14.44
14.50
14.61
14.64
14.70
14.73
14.74
14.81
14.86
15.04
15.11
15.15
15.60
15.83
16.00
16.20
104
-------
INTER-LABORATORV STUDY
POLLUTANT - N03
SAMPLE NUKBER - 4
M 48
TRUE-YALUE 10.50
SEAN 10.33
MEDIAN 10.40
DATA IN ASCENDING ORDER
378
UNITS - HICR06RAHS PER CyBlC METER
RANGE
VARIANCE
Sift. OEV.
COEF. VAR
7.68
1.4S
1.2?
11.77
C.I.CUPPER> 1Q.67
C.I. (LOWE R) 9.98
SKEWNESS -.63
ACCURACY -.95
5.Co
5.98
7.77
8.56
8.61
8.97
9.03
9.18
9.41
9.48
9.72
9.82
9.83
9.86
9.89
9.90
9.92
10.02
10.20
10.24
10.26
10.30
10.37
10.39
10.40
10.40
10.48
10.50
10.50
10.50
10.56
10.59
10.60
10.61
10.67
10.76
10.82
11-00
11.06
11.16
11.2?
11.30
11.55
11.70
11.70
11.93
12.00
12.22
13.66
105
-------
INTER-LABORATORY STUDY 278
POLLUTANT - N03
SAMPLE NUMBER - 5
« 49
TRUE-VALUE 12.00
1E»N 11.57
MEDIAN 11.81 COEF. VAR
DATA IN ASCENDING ORDER
UNITS - HICROGR*«S PER CUBIC METER
RANGE
VARIANCE
12.32
3.56
1.89
16.31
C.I.(UPPER) 12.10
C.I.(LOWE R) 11.04
SKEyKESS -1.97
ACCURACY -1.58
3.28
5.58®
6.65
7.84
8.79
9.83
9.97
10.07
10.63
11.00
11.11
11.40
11.44
11.49
11.50
11.56
11.65
11.70
11.70
11.70
11. 72
11.73
11.76
11.79
11.80
11.81
11.83
11.90
11.94
11.94
11.95
11.98
12-00
12.10
12.13
12.15
12.17
12.18
12.22
12.43
12.48
12.54
12.77
12.80
12-89
13.02
13.12
14.28
14.45
15.60
19.680
106
-------
INTER-LABORATORY STUDy
POLLUTANT - N03
SAMPLE NUHBER - 6
II
TRUE-VALUE
HEAN
PtEDIAN
28
.00
.57
.24
DATA IN ASCENDING ORDER
278
UNITS - niCROGRAHS PER CuBIC METER
RANGE
VARIANCE
STD. pEtf.
COEf . VAR
8.99
2.77
1.66
290.76
C.I.
-------
INTER-LABORATORY STUDY
POLLUTANT - N02
SAKPLE NUMBER - 1
TRUE-VALUE
82
.13
.13
.13
DATA IN ASCENDING ORDER
678
UNITS
RANGE
VARIANCE
STO. DEV.
COEF.
BICRQ6RAPIS PER CUBIC NETER
.15
.00
.02
14.10
C.I.(WPPER) .13
C.I.CLOWE R) .1!
SKEWNESS 3.55
ACCURACY 2.38
• 09 •
.10
.10
.10
.11
.11
.11
.11
.11
.11
.11
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.12
.13
.13
.13
.13
.13
.13
.13
.13
.13
.13
.13
.13
.12
.13
.13
.13
.12
.13
.13
.13
.13
.13
.15
.13
13
.13
,13
,13
,13
,13
,13
,13
,14
,14
,14
,14
,14
,14
,14
,14
,14
,14
.14
.14
.14
.15
.15
.15
.15
.15
.17
.18
.24f
.25
.60§
6.450
108
-------
INTER-tABORAToRY STUDY
POLLUTANT - N02
SAMPLE NUKBER - Z
ft 84
.24
.25
•2S
678
MEAN
IE&1AN
UNITS - HICR06RAMS PER CUBIC MEyER
RANGE
VARIANCE
STD. tEV.
COEF. VAR.
.18
.00
.02
8.90
C.I.tUPPER) .26
C.I.(LOWE R) .25
S 1.58
Y 5.04
DATA IN ASCENDING ORDER
.!&•
.19
.20
.21
.22
.22
.22
.22
.22
.22
.23
.23
.23
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.27
.27
.27
.27
.27
.28
.28
.28
.26
.30 ^
.30
.30
.37
.«1*
12.09$
109
-------
INTEH-LABORAToRt STUDY
POLLUTANT - N02
SAWPLE NUMBER - 3
N
678
TRUE-VALUE
1ED1AH
DATA IN ASCENDING ORDER
84
.24
.25
.25
UMITS ~ KICROGRANS PER CUBIC METER
RANGE .16
VARIANCE .00
STt>. 0EV. .02
COEF. WAR. 8*61
C.I.(UPPER) .26
C.I.CLOWE R) .25
SKEUNESS 1.96
ACCUR^CV 4.58
.180
.21
.21
.22
.22
.22
.22
.22
.23
.23
.23
.25
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.24
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.26
.27
.27
.27
.27
.27
.27
.27
.27
.27
.28
.28
.29
.JO
.30
.33f
.37
.43ft
1.730
11.99ft
110
-------
INTER-LABORATORY STUDY
POLLUTANT - M02
SAMPLE NUMBER - 4
TRUE-VALUE
HEAN
HEOIAH
83
.37
.39
.39
MTA IN ASCENDING ORDER
678
UIUTS - M1CROGRAMS PER CUBIC METER
RANGE
VARIANCE
ST0. &Etf.
COEF. VAR
.IS
.00
.02
6.28
C.I.(UPPER) .40
C.I. (LOWE R) .3?
SKEyNESS 1.28
ACCURACY 4.Q2
.32
.35
.35
.35
.36
.36
.36
.36
.36
.36
.36
.36
.37
.37
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.38
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.39
.40
.40
.40
-40
•40
• 40
.40
.40
.40
.40
-40
-40
-40
-40
.41
.41
.41
.41
.41
.41
.4t
.41
.42
.42
.42
.43
.48
.50
.52f
:IX
i.oof
18.660
in
-------
INTER-LABORATORY STUDY
POLLUTANT - N02
SAMPLE NUMBER - 5
TRUE-VALUE
1EAN
MEDIAN
82
.69
.71
DATA IN ASCENDING ORDER
678
UNITS - RICR06RA«S PER CUBIC METER
RANGE .43
VARIANCE .00
STD* DEV. .05
COEF. VAR. 7.03
C.I.(UPPER) .72
C.I. (LOWE «) .69
SKEUNESS -3.34
ACCURACY 3.05
.120
.120
.39
.49t
.55
.63
.64
.66
.67
.67
.67
.67
.56
.63
.68
.68
.68
.68
.69
.69
.69
.69
.69
.69
.69
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
,70
,71
,71
,71
,71
,71
,71
,71
,71
,71
.71
,71
,71
,71
,72
,72
.72
,72
.72
.72
.72
.72
.72
.72
.72
.72
.72
,72
.73
.73
.73
.73
.73
.73
.73
.73
.74
.74
.74
.74
.75
.75
.75
.76
.77
.77
.77
.78
.82
.97 f
34.13 f
112
-------
INTER-LABORATORY STU0Y
POLLUTANT - N02
SAMPLE NUMBER - 1
TRUE-VALUE
1EAN
74
.26
.27
.27
DATA IN ASCENDING ORDER
1278
UNITS - HICROGRAMS PER CUBIC
RANGE
VARIANCE
STO. OEV.
COEF. VAR
.13
•00
.03
9.21
C.I.CUPPER) .28
C.I.CLOWE R) .27
SKEWNESS .73
ACCURACY 2.65
.23
.23
.23
.23
.23
.24
.24
.24
.24
.24
.25
.25
.25
.25
.25
.25
.25
.25
.25
.25
.26
.26
.26
.26
.26
.26
.26
.26
.26
.27
.27
.27
.27
.2?
.2?
.27
.27
.27
.27
.27
.27
.27
.27
.28
.28
.28
.28
.26
.28
.28
.28
.28
.28
.29
.29
.29
.29
.29
.29
.29
.29
.29
.29
•29
.30
.30
.30
.31
.31
.32
.33
.34
.36
.38
.59
.74
10.88
•
113
-------
INTER-LABORATORY STUDY
POLLUTANT - N02
SAMPLE NUMBER - 2
TRUE-VALUE
MEAN
fED1AN
74
.39
.39
.39
DATA IN ASCENDING ORDER
1278
UNITS - MICR06RANS P£R CUBIC METER
RANGE .13
VARIANCE .00
STD. OEV. .03
COEF. VAR. 6.62
C.I.(UPPER) .40
C.I.CLOWE R) .39
SKEWNESS .01
ACCURACY .52
.33
.34
.34
.34
.34
.35
.36
.36
.36
.37
.37
.37
.37
.37
.37
.37
,38
,3E
,38
,38
,38
,38
,36
,38
,38
.38
.38
,38
.39
,39
,39
,39
.39
.39
.39
.39
.39
.39
.39
.3V
.39
.39
.39
.40
.40
.40
.40
.40
.40
.40
.41
.41
.41
.41
.41
.41
.41
.41
.4t
.41
.41
.41
.42
.43
.43
.43
.43
.43
.43
.43
.45
.46
.56
.SO
.88
15.90
*
f
f
114
-------
INTER-LABORATORY STUDY
POLLUTANT - N02
SAMPLE NUMBER - 3
TRUE-VALUE
HE AN
REDIftN
71
.52
.52
.52
DATA IN ASCENDING ORDER
1278
UNITS - PUCROGRAMS PER CUBIC HETER
RANGE
VARIANCE
STD. DEV.
COEF. VAR
.22
.00
.04
7.57
C.I.(UPPER) .53
C.I.CLOWE R> .51
SKEHNESS -.63
ACCURACY .78
.39
.41
.41
.46
.46
.47
.47
.47
.48
.49
.49
.49
.49
.49
.49
.50
.50
.50
.50
.50
.50
.51
.51
.51
.51
.51
.51
.51
.52
.52
.52
.52
,52
,52
.52
,51
,52
,52
-52
,52
,52
,52
,53
,53
,53
,53
,52
.53
.54
.54
.54
.54
.54
.54
.54
.54
.55
.55
.55
.55
.55
.55
.55
.56
.58
.58
.58
.59
.59
.61
.72
.89
.97
20.91
•
115
-------
INTER-LABORATORY STUB*
POLLUTANT - N02
SAMPLE NUMBER - 4
1278
TRUE-VALUE
MEDIAN
DATA IN ASCENDING ORDER
74
.70
.71
.70
UNITS ~ WICR06RAMS PER CUP1C METER
C.J.(UPPER) .71
C.I.CLOWE R) .70
SKEWNES5 .61
ACCURACY -.00
RANGE
VARIAWCE
STD. DEV.
COEF. VAR .
.26
.04
5.83
.63
.63
.65
.65
.65
.67
.67
.67
.67
.67
.68
.68
.68
.68
.68
.68
.68
.68
.69
.69
.69
.69
.69
.69
.69
.69
.69
.69
.69
.69
.69
.71
.70
.70
.70
.70
.70
.70
.70
.70
.71
.71
.71
.71
.71
.71
.71
,71
,72
,72
72
72
,72
,72
.72
,72
,72
,73
,73
,73
,73
,73
,73
,74
,74
,74
,76
,77
.78
.80
,81
.82
.84
.03
,14
,29
28.03
116
-------
INTER-LABORATORY STUDY
POLLUTANT - N02
SAMPLE NUMBER - 5
I
1278
TRUE-VALUE
JEOIAN
DATA IN ASCENDING ORDER
73
.93
.94
.94
UNITS - MICR06RAHS PER CUBIC METER
RANGE .74
VARIANCE .01
STO. OEV. .08
COEf. VAR. 8.38
C.I.(UPPER) .96
C.I.(LOWE R) .93
SKEWNESS -2.75
ACCURACY .97
.45
.80
.83
.85
.87
.S8
.89
.89
.90
.90
.90
.90
.91
.91
.91
.91
.92
.92
.92
.92
.92
.92
.92
.92
.92
.92
.93
.93
.93
.93
.93
.93
.94
.94
.94
.94
.94
.94
.94
.95
.95
.95
.95
.95
.95
.95
.95
.96
,96
,96
,96
,96
,96
,96
,96
,97
,97
,97
,97
,97
,97
,98
.98
,99
1.00
1.00
1.00
1.03
1.03
1.04
1.04
1.07
1.19
1.34 •
1.40
1.68
37.23 •
v
117
-------
!ME K-L A[ Of-'fc TGPY STUfY
TOLL U ANT - CC
CA*PLL M.MPLF - 1
1*2
THUL-VALUE 7.15
-EAN 7.01
yici «N 7.cc
DATA IN ASCENDING ORDER
3 78
UK ITS - HI CPCGRAMS PET CUDK M F T£ z
PANGF
VARIANCE
S T 0. CE V .
COEF.
C.I.1UFPLP ) 7.2?
C.J.ILIWE R) fc.EC
SKEWNESS lb.36
ACCURACY -2.10
• 21§6. 26
4.00 6.30
4.30 6.30
4.40 6.30
4.50 6.30
4. BO 6.34
5.00 6.35
5.00 6.36
5.00 6.40
5. DO 6.40
5.32 6.40
5.46 6.40
5.50 6.41
5.50 6.47
5.50 6.50
5.50 6.50
5.50 6.50
5.60 6.50
5.61 6.50
5.69 6.50
5.80+ 6.50
5.80 6.50
5pfc f^
• BO
5r* f^.
• BC
5n r
.95
6«h A
.00
6f* f^
.00
6.00
6.00
6.00
6. DO
6.00
6.07
6.10
6.10
6.10
6.20
6*% M
• 20
6.20
6.20
6.22
6.25
6.25
6.50
6.50
6.50
6.5C
6.50
6.5C
6.50
6.50
6.50
6.50
6. 50
6.50
6.55
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.60
6.63
6.65
6.65
6.69
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.70
6.74
6.75
6.75
6.75
6.77
6.80
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
.80
.80
.80
.80
.80
.80
.80
.80
.80
.80
.80
.80
.80
.81
.85
.87
.88
.89
.90
.90
.90
.90
6.90
6.9c
6. 90
6.90
6.90
6.9C
6. 91
6.9!
6.93
6.9?
6.94
6.95
6.96
6.96
6.99
7.00
7.0C
7. 00
7.00
7.00
7. 00
7. 00
7.00
7.00
7.00
7.00
7.00
7.00
7.0C
7.00
7.00
7.00
7.0C
7.00
7.00
7.00
7.0C
7.00
7.0C
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7. or
7.00
7.00
7.0C
7.0U
7.00
7.0C
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7.00
7-00
7-00
7-00
7.00
7.0o
7.QO
7-00
7-00
7« 00
7.Q4
7.Q 5
7.Q9
7.1o
7.1o
7.10
7.10
7.1Q
7.1o
7.1o
7.14
7.19
7.20
7.20
7.20
7.20
^ " W
7.20
7.20
~* w
7.20
^ ^i* ^j
7.20
7.2o
7.20
7.2o
7.2Q
7.2o
7.2Q
7.22
7.25
7.25
7.25
7.25
7.25
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
.28
.28
•30
.30
•30
.30
-30
-30
.30
.33
.35
.39
-40
• 40
-40
•40
.40
-40
.43
.46
•50
•50
u
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.50
7.60
7.60
7.60
7.70
7.70
9 V 9 Igf
7.70
7.70
7.72
7.75
f * r J
7.75
. v • ^
7. 80
• • \i U
7.80
' • *J \J
7.80
• 9 \*f ^j
7.80
7.82
7.88
7.90
7.90
7.91
" W '" W
f nn
c *u u
£.00
8 ~nn
. w u
s.nn
w * u u
8nn
• u u
£.00
8.00
8.00
8.00
8.05
8.05
8.10
8.20f
8.26
8.26
8.30
8. 50
8.50
8.50
9.00
9.00
9.00
9.00
10.10
14.101
38.00|
46.90
118
-------
1NTL--L At'ORA TORY STUTY
POLLITANT - CO
SAHPIE NUMBEP - 2
UNITS - HICPOCRAMS PEP CUBIC METER
TKUE-VALUE 2D.?3
HEAN 2H.33
KEPISN 2°. so
BAT* IN ASCENDING OR&ER
RANGE
VARIANCE
STP.
16.00019.50
16.20 19.53
16.BO 19.60
17.00 19.60
17.39 19.60
17.50 19.67
17.60 19.70
17.70 19.70
17.80 19.72
17.83 19.75
18.00 19.75
18.00 19.75
18.20 19.80
18.30 19.80
18.72 19*80
18.75 19.80
18.80 19.90
18.90 19.90
18.93 19.90
19.00 19.90
19.00 19.94
19.00 20.00
19.00
19.00
19.10
19.10
19.20
19.20
19.20
19.20
19.20
19.30
19.40
19.40
19.40
19.40
19.50
19.50
19.50
19.50
19.50
19.50
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20. 00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.04
20.05
20.10
20.10
20.10
20.10
20.10
20.10
20.10
20.10
20.10
20.13
20.14
20.15
20.IB
20.20
20.20
20.20
20.20
20.20
20.20
20.20
20.20
E
V.
AK.
20. 2C
ZD.ZC
20.20
20.21
20.21
20.22
20.24
20.25
20.25
20.25
20.25
20.25
20.25
20.27
20.30
20.30
20.30
2D.3C
20.30
20.30
20.30
20.30
20.30
20.30
20.30
20.30
20.30
20.30
20. 3C
20.30
20.30
20.30
20.35
20. 3S
20.39
20. 40
20.40
20.40
20.40
20.40
20.40
20.40
20.40
119
7.2D
.96
.96
M.g]
20.40
20.40
20.40
20.40
20. 40
20. 40
20.47
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.50
20.5p
20.50
•20.5o
20.50
C
C
.1.IUFPE*
. I. ILCklE
<• ) ?C .
R ) 2
no
G. 23
SKEWMESS -.73
A
20.50
20-50
20.50
20.50
20. 50
20. 50
20.5o
20.50
20. 50
20. 50
20. 50
20. 50
20.50
20. 50
20.54
20.55
20.56
20.6o
20.60
20.60
20.60
20.60
20.60
20.60
20.60
20.65
20.70
20.70
20.7Q
20.7Q
20.70
20.7o
20.70
20.71
20.71
20.72
20.72
20.72
20.73
20.75
20.80
20.8o
20.80
CCURACY
20. SO
20.80
20.8Q
20.8o
20.8Q
20.82
20. 90
20.90
20. 90
20.96
20.98
21.QO
21-QO
21-QO
21.QO
21.QO
21-00
21.00
21-00
21.00
21.00
?1.00
21.00
•
21.00
21.00
21.00
21.00
21.00
21.00
21.00
21.00
21.00
21.00
21.10
21.13
21.20
21.25
21.30
21.30
21.30
21.30
21.31
21.34
21.40
21.40
21.48
21.50
21.50
21.50
21.50
21.60
21.60
21.60
21.60
21.64
21.70
21.70
21.70
21.70
21.70
21.70
21.70
21.70
21.70
21.74
21.75
35
21.75
21.79
21.80
21.80
21.94
22.00
22.00
22.00
22.05
22.19
22.20
22.70
22.78
22.80
22.89
23.40
23.500
40.200
-------
IKTE.c-lAfGFATCRY STL; TV
^'GLLUtMT - CC
:?£
UNITS - MICROGRAMS PEF CUPIC
3 1C
HUf_ -VALUE
-L'^N
"LC I :r
<*2. ?-fc
12. 2G
«?. 3C
DATA IN ASCENDING
.67
4.75
5.00
6.30
33.80
34.20
35.65
37.43
38.40
38.50
38.70
39.00
39.00
39.00
39.40
39.40
39.50
39.50
39.88
39.90
40.00
40.00
40.00
40.00
40.00
40.00
40.50
40.50
40.50
40.70
40.70
40.75
40.50
40.90
41.00
41.00
41.00
41.00
41.00
41.00
41.00
41.00
41.00
®41. 06
41.10
041.15
41.20
41.20
041.20
41.20
41.25
41.30
41.30
41.37
41.40
41.43
41.46
41.50
41.50
41.50
41.50
41.50
41.50
41.50
41.50
ORDER
41.50
41. 5D
41.50
41.50
41.50
41.50
41.50
41.50
41.60
41.60
41.60
41. 70
41.70
41.75
41.75
41.75
41.79
41.80
41.80
41.80
41.80
41.80
41.84
41.88
41.90
41.90
41.90
41.90
41.90
41.90
41.90
41.93
41.97
42.00
42.00
42.00
42.00
42.00
42.00
42.00
42.00
42.00
42.00
VAR] AfcC
STD. CE
COEF. V
42.00
.42.00
42.00
42.00
42.00
42.00
42.00
42.00
42.00
42.00-
42.00
42.00
42.00
42. OQ
42.00
42.01
42.03
42.05
42.10
42.10
42.10
42.10
E
V.
AP.
42.10
42.10
42.1C«
42.14
42.16
42. 1«
42. 1£
42.20
42.20
42.20
42.20
42. 2C
42.20
42.20
42.20
S2.2C
42.25
42.25
42.25
42.25
42.29
42.30
42.30
42.30
42.30
42.30
42.30
42.30
42.30
42.30
42.30
62.30
42.33
42.35
42.36
42.3V
42.40
42.40
42.40
42. 4C
42. 4C
42. 4C
42. 4C
1 20
10.17
3. 19
7.56
42.40
42.40
42.44
42.45
42.46
42.46
42.50
42.50
42.50.
42.50.
42.50
42.50-
42.50
42.50
42.50
42. 50
42. 50
42.50
42. 5C
42.50
42.50
42. SQ
C.I.fLCUE
SPEWNFSS
ACCURACY
KJ
42-50 43.00
42.50 43.00
42.59 43.QO
42.60 43.QO
42.6Q 43.0o
42.60 43.OQ
42.62 43.oo
42.64 43.20
42.7Q 43.25
42.70 43.3o
42.70 43.3o
42.70 43.3Q
42.70 43.4o
42.7Q 43.4Q
42.70 43.4o
42.7o 43.4Q
42.72 43.42
42.75 43.42
42.8Q 43.50
42.&Q 43.5o
42.8Q 43.50
42.80 43.50
42.80
42.SO
42.80
42.80
42.8Q
42.50
42.80
42.81
42.83
42,90
42.90
42.99
43.QO
4-3,00
43.QO
43.00
43.QO
43.00
43.QO
43.QO
43.nn
43.50 -45.10
43.50 45.20
43.50 45.30
43.50 45.31
43.50 45.40
43.50 45.40
43.50 45.50
43.50 45.50
43.60 45.60
43.60 45.70
43.60 45.80
43.60 46.49
43.60 46.50
43.64 46.50$
43.70 46.70
43.70 47.02
43.80 47.25
43.80 47.40
43.80 84.30f
43.83
43.86
43.9t
44.00
44.00
44.00
44.00
44,00
44.02
44.04
44.10
44.10
44.10
44.15
44.16
44.18
44.30
44.40
44.53
44.60
44.70
44.80
45.00
45.00
-------
JMEC-L APOK ATGfcY STUDY
POILUANT - CO
978
UNITS - HICROGRAHS PE'1 CUBIC ("HE
N
TRUE
MEAN
HEDI
DATA
-VALUE
AN
331
7.. 92
3.69
3.60
RANGE
VAPIANCE
STD. CFV.
COEF. VAR
*
£ . 50
W V w* 1— J
.62
. PC
• S-L \*J
21.51
III ASCENDING ORDER
1.5Q
1.90
2.900
2.10
2.15
2.20
2.29
2.500
2.50
2.50
2.50
2.50
2.60
2.60
2.60
2.60
2.60
2.610
2.62
2.65
2.67
2.70
2.75
2.79
2.80
2.80
2. SO
2.84
2.87
2.90
2.90
2.90
2.90
2.960
3. DO
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.01
3.04
3.04
3.04
3.04
3.10
3.10
3.10
3.10
3.10
3.10
3.10
3.10
3.12
3.13
3.13
3.15
3.19
3.19
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.22
3.25
3.25
3.25
3.25
3.29
3.30
3.30
3.30
3.30
3.30
3.30
3.30
3.30
3.30
3.31
3.33
3.34
3.35
3.35
3.36
3.37
3.38
3.40
3.40
3.40
3.40
3.40
3.40
3.40
3.42
3.42
3.42
3.43
3.44
3.44
3.45
3.45
3.45
3.48
3.49
3.50
3. 50
3.50
3.50
3.50
3. 50
3.50
3.50
3.50
3.50
3.50
3.50
3. 50
3.50
3. 50
3.50
3.50
3.50
3.56
3.50
3.50
3.50
3.50
3.50
3.50
3.50
3.50
3.50
3.51
3.52
3.53
3.55
3.56
3.57
3.57
3.57
3.59
3.60
3.60
3.60
3. 60
3.60
3.60
3.60
3.6C
3.6G
3.60
3.60
3.60
3.60
3.62
3.64
3.65
3.66
3.66
3.6?
3.7C
3.7C
3.70
3.70
3.70
121
3.70
3.7C
3.7C
3.70
3.70
3. 70
3.70
3.74
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.75
3.77
3.79
3.80
3.80
3.8G
3.8D
3.8C
3.80
3.8o
3.80
C.I.CLCU-E
SKEWNESS
ACCIFP/!CV
3.7P
RJ
3.SO
3.80
3.8o
3.81
3.86
3.89
3.90
3.9Q
3.90
3.9Q
3.9Q
3.9o
3.9o
3.9Q
3,
3,
3,
3,
3,
.90
,90
•90
.90
.90
3.91
3.91
3.93
3.94
3.95
3.99
4.00
4.00
4-00
4.00
4.QO
4-00
4.00
4-00
4.QO
4.QO
4-00
4.00
4-00
4-00
4-00
4.QO
4-00
4«QO
4.00
4.00
4*00
4-00
4.00
4.00
} 4.QO
4-00
4-00
4-00
4.QO
4. QO
4-00
4.00
4«00
4-00
4-00
4-00
4-00
4-00
4.QO
4-00
4-00
4-00
4.00
4.00
2 .77
-5 .76
4.00
4.00
4.00
4.01
4.08
4.10
4.10
4.10
4.10
4.13
4.16
4.17
4.20
4.20
4.20
4.20
4.20
4.20
4.20
4.20
4.25
4.25
4.25
4.29
4.30
4.30
4.30
4.30
4.35
4.35
4.35
4.38
4.40
4.40
4.43
4.50
4.50
4.50
4.50
4.50
4.50
4.50
4.58
4.70
5.00
5.00
5.00
5.00
5.10
5.10
5i,62
5.88
6.000
6.00
6.00
6.00
6.30
6.90
7.000
7.390
7.6Q0
7.710
7.83
8. 00
10.00
11.000
-------
TKTE r-L A<"-0& ATOPY S TUrY
rCLLlTA\T - CC
'AfKLi MMBER - :
UNITS - MI CROC RAM 5
CUP 1C
N
UU£ -VA
ME* N
*LCI «N
DATA IN
8
9
11
11
11
12
12
12
12
12
12
12
12
12
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
13
Lit
3 3
1^.5
m • P
11."
ASCENDING
.510
.1 00
.000
.30
.740
.07
.090
.43
.50
.78
.79
• eo
.90
.90
.00
.11
.20
.20
.25
.29
.30
.30
.34
.39
.39
.40
.40
.40
.48
.50
.50
.50
.50
.58
.60
.60
.50
.60
.50
.60
.65
.66
.70
13
13
13
13
13
13
13
13
13
13
14
14
14
14
14
14
14
14
14
14
14
14
.75
.50
.?0
.80
.80
.50
.84
.90
.91
.99
.00
.30
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
14.30
14
.05
14.07
14
.10
1
fc
n
0
RANGE:
VARI 6NC
STD. DC
COEF. V
E
V.
A? .
32. 1C
3.27
1.61
12.18
C .
c.
SK
I . Ill
I . (I
EWNE
A CCU& A
IFP.
ow
s s
CY
ORDER
14.10
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
14
.10
.10
.11
.12
.13
.17
14.50
14.50
14.50
14.50
14.50
14.50
14.50
.20014.50
.20
.20
.20
.20
.20
.25
.25
.25
.25
.25
.25
.30
.30
.30
.30
.30
.30
.30
.30
.31
.38
.38
.40
.40
.40
.43
.45
.46
.48
.49
.50
.50
.50
.50
.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.50
14.52
14.56
14.56
14. 5B
14.59
14.60
14.60
14.60
14.60
14.60
14.60
14.60
14. 6Z
14.63
14.65
14.66
14.66
14.67
14.69
14.70
14.70
14.70
14.70
14.70
14.70
14.73
14.74
14.75
14.75
14.75
14.80
14.80
14.80
14.80
14.80
14.80
14.80
14.80
14.80
14.80
14.81
14.84
14.84
14.85
122
14.88
14.90
14.90
14.90
14.90
14. 9G
14.90
14.90
14.90
14. 9D
14.96
14.99
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15. Oo
15.00
15.00
15.QO
15.00
15.QO
15.QO
15.QO
15.QO
15.QO
15.QO
15.QO
15.00
15.QO
15.QO
15.QO
15.00
15.00
15.QO
15.00
15.QO
15.QO
15.00
15.00
15.QO
15. oo
15.QO
15.QO
15. 01
15.05
15. 0^
15.10
15.10
15.10
15.1o
15.12
15.13
15.13
15.20
15.20
15.2o
15.20
15.20
15.20
15.
20
15.20
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
15.
eo
22
23
30
30
30
30
30
30
32
33
40
*0
«0
40
*0
50
50
50
15.50
15.
15.
15.
•50
.50
,50
15.50
I 15.U3
Rl 14.65
11 .92
1 .65
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.50
15.52
15.60
15.60
15.60
15.60
15.66
15.70
15.70
15.70
15.70
15.75
15.80
15.80
15.80
15.80
15.80
15.80
15.80
15.90
15.95
16.00
16.00
16.00
16.00
16.00
16.10
16.10
16.17
16. 24
16.30
16.30
16.30
16.40
16.50
16.52
16.60
16.70
17
17
17
17
18
19
19
19
19
00
00
00
08
00
00
ool
00
00
20.62|
21.31J
22.71
32.10|
43.40
«
-------
P-l AFOR ATGRY STUPY
POLL n ANT - CO
SAMPLE NljHBER - 3
UNITS - KICPOfRAMS PEP CUBiC KETE?
N
TRUE-VA
ME»N
HEBI*N
DATA IN
16
22
30
31
31
32
32
LUE
3 71
36. MS
36. 9t
36.92
RANCE
VARIANCE
STD. CEV.
COFF. VA5
.
25.5C
2.95
1.72
««.fc5
C.
C.
I . ( u r- P
I. ILCW
SMEViNFSS
ACCURACY
ASCENDING ORDER
.50
.600
.520
.000
.830
35.55
35.57
35.60
35.63
35.70
• 00035.72
.65
33.40
33
.74
33.91
34
34
34
34
34
34
34
34
34
35
35
35
35
35
35
35
35
35
35
35
35
.44
.50
.50
.50
.50
.50
.61
.70
.70
.00
.00
.00
.10
.20
.20
.27
.30
.40
.40
.40
.42
35.72
35.80
35.85
36.000
36.00
36.00
36.00
36.00
36.00
36.00
36.00
36.00
36.00
36.00
36. DO
36.00
36.00
36.00
36.00
36.00
35.44
35
35
35
35
35
35
35
35
35
35
35
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.55
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
00
06
06
09
09
10
10
10
ID
10
15
15
20
20
20
20
20
25
36. 28
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
36.
30
30
30
31
33
39
40
40
40
40
40
42
49
50
50
50
50
50
50
50
50
50
5D
50
36.50
36.50
36.50
36.50
36.50
36. 5D
36.50
36.50
36.50
36.50
36.50
36.50
36.51
36.54
36.54
36.54
36.55
36.56
36.56
36.60
36.60
36.60
36.60
36.63
36.63
36.70
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
37
37
37
*
*
•
*
*
»
»
*
•
*
•
•
•
•
•
*
*
.
*
•
•
*
*
*
*
*
*
*
•
»
m
*
»
•
*
*
*
*
37.
37
37
37
37
*
*
*
*
1
70
70
70
70
70
70
70
70
70
70
70
72
75
75
80
80
80
80
80
80
86
87
gg
89
90
90
90
9C
90
90
90
92
92
95
99
OC
00
00
00
00
00
00
CO
23
•37.00
37.00
37.00
37.00
37.00
37. DG
37. OL
37.00
37.00
37.00
37.00
37.00
37.00
37.00
37.00
37.00
37.00
37.00
37.01
37.04
37.05
37.10
37.10
37.10
37.16
37.16
37.20
37.20
37.20
37. 20
37. 20
37.20
37.25
37.25
37.25
37.25
37. 30
37. 30
37.30
37.3o
37.30
37.30
37.35
37.35
37.40
37.40
37.4o
37.4Q
37.40
37.4Q
37.40
37.40
37.4Q
37.40
37.44
37.50
37.SQ
37.50
37.50
37.50
37.5Q
37.50
37.50
37.50
37.5o
37.50
37.50
37.5Q
37.50
37.52
37.55
37.60
37.60
37.6Q
37.62
37.65
37.7o
37.7Q
37.7o
37.75
37. BO
37.8Q
37.80
37.80
37.83
37.90
38.0o
38.QO
38.QO
38.QO
38. QO
38.QO
38. oo
38.QO
38.00
-
} 3 7 .1 ft
P ) 36 . 77
-U .83
i .:
-------
INTER-LABORATORY STUDt
POLLUTANT - PB
SAMPLE NUMBER - 1
* 61
TRUE-VALUE 6.<>0
1EAN 6.40
"EDIAN 6.42
DATA IN ASCEMDIN6 ORDER
178
UNITS - HICR06RAMS PER CUBIC METER
RANGE
VARIANCE
ST&. DEtf.
COEf. VAR
7.32
.77
.86
13.69
C.I.(UPPER! 6.6?
C.I.CLOWE R> 6.18
SKEWNESS -2.74
ACCURACY -2.73
1.38
1.80
4.37
5.08
5.10
5.39
5.40
5.51
5.59
5.91
6.00
6.00
6.12
6.12
6.13
6.13
6.14
6.17
6.20
6.23
6.23
6.26
6.30
6.30
6.30
6.30
6.32
6.36
6.36
6.36
6.36
6.3E
6.41
6.42
6.45
6.5G
6.52
6.54
6.56
6.60
6.60
6.66
6.69
6.69
6.7Q
6.70
6.72
6.74
6.75
6.76
6.78
6.84
6.85
6.9Q
6.92
7.10
7.11
7.13
7.28
7.46
£.10
8.70
9.22
22.61
39.69§
124
-------
INTER-LABORATORY STUD*
POLLUTANT - P8
SAMPLE NUMBER - 2
* 6t
TRUE-VALUE 12.60
iEAN 11.92
1EDIAN 12.12
DATA IN ASCENDING ORDER
178
UNITS - BICR06RAHS PER CUBIC METER
RANGE
VARIANCE
STD. OEV.
COEF. VAR
14.76
3.15
1.77
14.89
C.I.(UPPER) 12.36
C.I.CLOWE R> 11.47
SjCEWNESS -2.68
ACCURACY -3*81
2.33
3.320
6.36
8.58 •
8.62 •
9.06
10.20
10.60
10.900
11.05
11.16
11.40
11.40
11.40
11.40
11.42
11.50
11.52
11.55
11.61
11.61
11.70
11.70
11.80
11.88
11.90
12. OB
12.00
12.00
12.00
12.00
12.00
12.04
12.07
12.12
12.14
12.2C
12.23
12.29
12.3o
12.3o
12.35
12.40
12.43
12.47
12.48
12.48
12.48
12.48
12.5Q
12.56
12.58
12.60
12.63
12.66
12.74
12-75
12.76
12.80
12.80
12.96
13.20
14.10
14.40
16.22
17.09
125
-------
INTER-LABORATORt STUDY
POLLUTANT - PB
SAHPLE NUMBER - 4
TRUE-VALUE
1EAN
1 ED IAN
E2
.00
.14
•Q5
DATA IN ASCENDING ORDER
17E
UNITS - KICROGRANS PER CUBIC METER
RANGE
VARIANCE
STD. DEW.
COEF. VAR
1.19
.0?
.26
180.59
C.I.(UPPER) .25
C.I.(LOME R) ,
SKEW NESS 3.12
ACCURACY .00
•31
.01
.01
.01
.01
,02
,02
.03
.03
,04
.05
• 06
.08
.10
.12
.15
,18
,19
,31
.42
1.20
18.66
126
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAWRLE NUMBER - 5
« 62
TRUE-VALUE 10.20
HEAN 9.72
10.04
DATA IN ASCENDING ORDER
178
UNITS - «1CROGRAHS PER CUBIC NETER
RANGE
VARIANCE
SID. DEV.
COEf .
13.35
3.25
t.g'D
18.56
C.I.CUPPER) 10.17
C.I.CLOWE R) 9.27
SKEUNESS -2.90
ACCURACY -1.57
•5?*
1.10
1.27 0
1.93
2.500
5.64
7.410
8.200
8.96
8.99
9.00
9.00
9.DO
9.06
9.10
9.120
9.34
9.36
9.36
9.42
9.42
9.48
9.50
9.60
9.60
9.60
9.62
9.70
9.78
9.B8
9.90
9.90
9.90
9.96
9.93
1D.OC
10.03
10.05
ID.06
10.OB
10.OB
ID.Ofc
10.09
10.10
10.10
10.16
10.20
10.2o
10.24
10.26
10.26
10.28
10.31
10.34
10.39
10.4Q
10.48
10.55
10.58
10.60
10.62
10.?0
10.80
10.80
11.20
11.20
11.65
13.44
14.45
127
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAMPLE NUMBER - 7
TRUE-VALUE
KEftN
tEDIAN
DATA IN ASCENDING ORDER
178
62
.60
.64
.60
RANGE
VARIANCE
SID. DEV.
COEf. VAR.
1.82
.05
.23
55.65
UNITS - NICR06RAMS PER CUBIC METER
C.I.(UPPER> .70
C.I.(LOWE R) .58
SKEWNESS 3.46
ACCURACY .00
.16
.20
.23
.38
.42
.42
.45
.46
.48
.50
.50
.50
.53
.54
.54
.54
.55
.56
.57
.57
.57
.58
.58
.58
.59
.59
.60
.60
.60
.60
.61
.60
.60
.6C
.60
.60
.60
.61
.61
.61
.62
.62
.62
.63
.63
.63
.63
.63
.64
.65
.66
.67
.67
.67
.67
.69
.70
.70
.72
.73
.79
.81
.86
.87
.94
1.29
1.98
1.980
2.97$
128
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAMPLE NURBER - 8
N
TRUE-VALUE
MEAN
MEDIAN
61
Z.04
1.98
2.04
DATA IN ASCENDING ORDER
178
UNITS - HICROGRAHS PER CUBIC METER
RANGE
VARIANCE
STD. t>EV.
COEF. VAR.
1.97
.10
.31
15.82
C.I.(UPPER) 2.06
C.I.CLOWE R) 1.90
SKEWNESS -3.22
ACCURACY .00
.43
.46 •
.63
1.45 +
1.50
1.70
1.740
1.75
1.770
1.80
1.BO
1.80
1.B3
1.86
1.66
1.66
1.86
1.92
1.92
1.93
1.95
1.96
1.96
1.97
1.98
1.98
1.98
2.00
2.00
2.00
2.02
2.02
2.04
2.04
2.04
2.04
2.04
2.04
2.04
2.05
2.06
2.8
2.08
2.08
2.09
2.1o
2. to
2. 1o
2.1Q
2.10
2.15
2.15
2.16
2.16
129
-------
INTER-LABORATORY STUDY
POLLUTANT - Pb
SAMPLE NUWBER - 9
N 64
TRUE-VALUE 1.50
"EAN 1.51
"ED1AN 1.52
DATA IN ASCENDING ORDER
678
UNITS - MICR06RAHS PER CUBIC METER
C.I.(UPPER) 1.57
C.I.CLOWE R) 1.46
SKEMNESS -.57
ACCURACY 1.00
RANGE
VARIANCE
STD. DEV.
COEF. VAR.
1.85
.05
.22
14.55
.4B
-94t
1.100
1.10
1.120
1.26
1.37
1.31
1.31
1.35
1.38
1.38
1.39
1.39
1.41
1.41
1.44
1.44
1.44
1.44
1.44
1.45
1.45
1.46
1.46
1.46
1.47
1.49
1.50
1.5o
1.50
1.50
1.50
1.51
1.51
1.52
1.52
1.53
1.52
1.55
1.55
1.55
1.56
1.56
1.56
1.56
1.56
1.57
1.58
1.58
1.58
1.59
1.6Q
1.60
1.61
1.61
1-62
1.62
1.62
1.6Z
1.63
1.65
1.6&
1.83
1.86
1.92
2.08
2.33
130
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAMPLE NUMBER - 0
678
UNITS - KICR06RAMS PER CUBIC METER
TRUE-VALUE
JEAN
HEDJAN
64
.42
.49
RANGE
VARIANCE
STO. DEV.
CtDEF. VAR.
4.43
.27
.52
105.93
DATA IN ASCENDING ORDER
C-I.COPPER) -61
C.I.CLOWE R) .36
SKEUNESS 7.29
ACCURACY -,00
• 06
.to
« •» A
.17 9
.29
.30
.30
.35
.36
.36
.370
.38
.39
.40
.40
.40
,.40
.40
.41
.41
.41
.41
.41
.41
.41
.42
.42
.42
.42
.42
.42
.42
.42
.42
.42
.42
.42
.43
.44
.44
.44
.44
.44
.44
.44
.44
.45
.45
.45
.45
.45
.45
.46
.46
.47
.47
.48
.48
.48
.48
.49
.54
.600
.60
v %^ \j
.60
.60
.61
1.800
4.49
131
-------
INTER-LABORATORt STUDT
POLLUTANT - PB
SAMPLE NUMBER - 4
N
TRUE-VALUE
JEAN
MEDIAN
15
.00
.74
.05
DATA IN ASCENDING ORDER
678
UNITS - HICR06RAMS PER CUBIC METER
RANGE
VARIANCE
SID. DEV.
COEF. VAR
9.71
6.20
2.4?
335.69
C.I.(UPPER) 2.00
C.I.CLOWE R) -.5?
SKEKNESS 3.10
ACCURACY .00
.01
.01
.01
01
,02
,03
,05
,05
>Q6
,12
.12
.15
.76
1.200
9.72
132
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAMPLE NUMBER - 5
678
N
TRUE-VALUE
WEAN
1EDIAN
64
9.90
9.80
9.90
UNITS - fllCROSRAMS PER CUBIC METER
&ATA IN ASCENOINS ORDER
RANGE
VARIANCE
ST0. DEV.
COEF. VAR.
13.12
3.78
1.94
19.82
C.I.COPPER) 10.28
C.I.(LOWE R) 9.3?
-1.52
1.S8
1.96
6.25
6.850
7.49
7.500
8.40
8.52
8.62
9. 00
9.00
9.00
9.00
9.06 '
9.12
9.14
9.30
9.38
9.57
9.59
9.60
9.61
9.63
9.64
9. 68
9.68
9.72
9.72
9.72
9.72
9.85
9.90
9.90
9.90
9.90
9.90
9.93
9.96
9.99
10.00
10.00
10.10
10.10
10.14
10.23
10.25
10.3Q
10.32
10.32
10.36
10.39
10.5o
10.5o
10.50
10.65
10.68
10.68
10.69
10.73
11.10
11.40
11.74
12.60
13.200
13.31
14.60
15.00
133
-------
INTER-LABORATORY STUDY
POLLUTANT - PB
SAMPLE NUMBER - 7
* 65
TRUE-VALUE 12.00
*E*N 11.58
MEDIAN 11.65
DATA IN ASCENDING ORDER
678
UNITS - HICR06RAHS PER CUBIC HETER
RANGE
VARIANCE
STB. DEV.
COEF. VAR
15.57
2.94
1.71
14.80
C.I.(UPPER) 11.99
C.I.(LOME R) 11.16
1t W r U U f t C inr-^ ft ft
ACCURACY -2.92
w
2.43
2.96$
8.26
8.40
8.35
9.17
9.49
9.60
9.BO
10.20
10.35
10.54
10.56
10.56
10.620
10.62
10.90
11.00
11.07
11.07
11.16
11.16
11.24
11.34
11.34
11.40
11.40
11.40
11.47
11.47
11.51
11.55
11.55
11.53
11.60
11.64
11.65
11.67
11.70
11.88
11.90
11.90
12.00
12.00
12.QO
12.QO
12.oo
12.QO
t2.03
12.Q5
12.09
12.15
12.16
12.27
12.28
12.47
12.47
12.57
12.70
12.72
12.72
12.75
12.78
12.80
12.PO
13.50
13.50
13.91
18.00
134
-------
INT t* -LABORATORY STUDY 1 Q78
POLLUTANT - S02 UNITS - HICRCGRAMS PER CUBIC METER
SAMPLE NUMBER - 1
L o ^l RAhlGE 42'45 C.I. (UPPER) 15.24
9.D7 VARIANCE 45.95 C.I. (LOWE R) 12.51
13't7 STf- OEV. 6.7B SKEtf*ESS 1.08
13.10 COEF. VAR. 48.88 ACCURACY 44.43
SATA IN ASCENDING ORDER
-55 10. 00 12.77 14.44 18.34
1-17 10.12 12.81 14.70 18.56
1-58 10.20 12.94 14. 90 18.70
2.62 10.40 13.100 15. QO 19.30
3.50 10. 4B 13.10 15.15 19.50
3.94 10.48 13.10 15.39 22.48
4-00 11.20 13.10 15.46 22.87
4.42 11.26 13.10 15.58 23.45
5.24 11.370 13.11 15.72 26.20
5.51 11.54 13.2fc 15.72 26.67
6.48 11.66 13.34 15.72 27.00
7.90 11.70 13.44 15.72 27.47
8.03 12.00 13.4fi 15.72 27.47
8.30 12.38 13.70 16. 2Q 28.77
9.40 12.53 14. 00 16.67 29.400
9.40 12.54 14.07 16. 7Q 30.40
10. DO 12.57 14.24 17.89 43.00
10.00 12.62 14.36 18.34
135
-------
1 NT ER-LA90R ATURV STUDY
PQLLJTANT - SG2
1C 78
SAMPLE
TRUE-VALUE
"EAN,
- 2
131
ib.EO
36.97
58.20
UNITS - MICROGRAMS PER CuPK METER
RANGE
VARIANCE
STO. DEV.
COEF. VAR.
61.56
1D1.02
10.05
27.19
C.I.(UPPER) 38.93
C.I.CLOWE R) 35.01
SKEWNESS -.01
ACCURACY -1.55
DATA IN ASCENDIfcS ORDER
8.9D
10.00
10.40
13.10
15.66
17.35
18.20
20.50
21.67
22.28
22.52
22.76
25.00
25.00
25.63
26.20
27.30
28.58
28.90
29.47
29.76
30.84
31.44
31.58
32.00
32.00
33.00
33.20
33.40
34.06
34.06
34.72
34.77
35.00
35.00
35.40
36.00
36.17
36.32
36.40
36.4£
36.5o
36.66
36.6S
36.68
36. 7C
37.11
37.30
37.90
38.00
38.26
38.30
38.34
38.41!
38.54
38.57
38.63
38.75
3B.83
38.86
38.98
39.50
39.30
39.30
39.3Q
39.3Q
40.50
40.50
40.86
40.93
40.95
41.67
41.92
41.92
42. Q?
42.48
42.70
43.69
44.00
44.00
44.23
44.54
44.54
45.30
45.55
46.00
46.15
48.33
49.78
51.55
54.00
60.80
63.33
70.46
97.40
136
-------
INTER-LABORATORY STUDY
POLLUTANT - SO?
SAMPLE NUMBER - 3
1
TRUE-VALUE
• EAN
HpnlAN
100
68 -.90
82.21
62.13
DATA IN ASCENDING ORDER
1078
UNITS - MICRCGRHMS PER CUF 1C METER
RANGE
VARIANCE
STD. DEV.
COEF. VAR
74.27
152.05
11.49
13.9B
C.I.(UPPER) 84.47
C.I.CLOWE R) 79.96
SKEWNESS -.56
ACCURACY -7.62
11.37
40.40
41.92
49.14
49.78
62.42
62.60
62.88
64.13
65.85
73.07
73.30
73.36
73.36
73.36
74.00
74.10
74.30
74.33
75.28
75.70
76.16
76.63
76.90
77.00
77.00
77.00
77.80
78.10
78.30
78.50
78.53
78.60
79.14
79.76
79.S3
80.14
80.30
80.40
eo.46
80.57
80.84
81.00
£1.20
81.2?
81.22
81.39
61.47
81.86
82. 30
82.33
82.46
B3.20
83.33
83.4C
83.7€
83.84
83.S4
84.00
85.QO
35.90
86.QO
86.25
86.46
86.46
86.61
86.70
86.81
36.97
87.37
87.42
88.QO
88.31
88.32
89.08
89.20
69.47
89.9&
89.99
90.06
90.75
91.41
91.70
91.70
91.70
91.76
96.15
100.20
102.18
104.17
105.30
110.00
110.00
164.40
137
-------
Y STUDY
1078
"OLLUTAM - S02
UNITS - MItROGRAMS PER
KETER
P - 4
S 99
TRUE-VALUE 1£8.!0
*EAN 1^5.66
MEDIAN 125.76
DATA IN ASCENDING ORDER
RANGE
VARIANCE
STD. DEW.
COEF. VAR
89.90
178-23
13.35
10.62
C.I.CLOWE
SKEWNESS
ACCURACY
128.29
R) 123.04
-.41
-1.9P
62.58(
78.60
79.BO
66.46|
93.01
99.00
99.36
103.50
103.87
106.59
110.04
110.04
111.10
112.58
112.66
113.50
113.97
114.00
116.10
117.00
117.70
117.80
119.00
119.28
120.00
tzo.to
120.96
121.00
121.17
121.49
121.50
121.82
122.72
122.86
123.14
123.t4
123.14
123.90
123.96
12*.01;
124.00
124.55
124.62
124.70
125.10
125.20
125.30
125.70
125.76
125.76
125.8^
126.50
126.90
1E7.4C
127.68
128.10
128.3fc
128.38
130.00
130.QO
130.Q8
130.21
130.3Q
130.81
131.05
131.55
131.60
132.32
132.5Q
132.7Q
132.85
133.33
133.62
133.62
133.62
133.88
134.90
135.54
135.77
136.24
136.24
136.78
137.00
138.08
138.46
139.82
142.31
144.65
145.00
149.40
151.96
165.00
250.50
138
-------
1NTER-LA90R ATORY STUDY
POLLUTANT - S02
SAMPLE NUMBER - 5
N 100
TRUE-VALUE 192.20
«E«N 168.91
"(EDIftN 169.45
DATA IN ASCEMCIMS ORDER
1078
RANGE
VARIANCE
SID. DEV.
COEF. VAR
- MICRCGR&MS PER CLINIC MITER
116.32
332.53
18.24
C.l.(UPPtR> 172.48
C.I.(LOWE R) 165.35
SKEMNESS -.09
ACCURACY -11.84
86.46 •
110.34
119.45
120.15
131.00
135.13
137.42
141.00
141.67
143.00
143.92
144.00
148.00
149.34
151.87
153.20
153.53
155.80
155.87
156.30
156.82
157.00
157.40
158.67
158.70
159.00
159.82
160.00
160.03
160.93
161.40
162.16
163.78
164.67
165.06
165.10
165.99
166.55
166.67
166.9C
167.15
16?.67
167.61
167.68
167.68
167.6I;
16P.OC
169.00
169.30
16-9.6(1
169.7C
170.60
170.62
172.7C
172.71
172.92
172.92
172.92
173.00
174.61
175.54
175.54
176.64
176.64
177.QO
178.16
178.4Q
179.60
179.7Q
179.98
180.55
180.74
180.91
182.07
182.29
183.00
183.40
183.48
183.70
184.01
184.16
185.00
1S6.00
186.02
187.89
189.20
190.61
194.70
203.85
206.60
209.60
226.66
274.80
139
-------
INTEK-LABOR ATOR Y STUDY
POLLUTANT - S02
SAMPLE NUMBER - 1
" 106
TRUE-VALUE 21.70
*EAN £0.75
"EDIAN 19.53
»ATA IN ASCENDING ORftER
478
UNITS - MICROGRAMS PER CUBIC METER
RANGE
VARIANCE
STO. OEV.
COEF. *AR
241.96
508.80
22.56
198.69
C.I.CUPPER) 25.05
C.I.CLOWE R) 16.46
SKCMNESS 9.08
ACCURACY -10.00
1.70
1.50 •
2.37
4.47
5.24
8.OOf
8.87 •
9.26.f
9.66
10.00
10.48
10.48
11.00
11.56
11.83
12.60
12.93
13.10
13.10
13.10
13.10
13.69
13.70
13.75
14. DO
14.00
14.90
14.93
15.00
15.00
15.15
15.29
15.39
15.72
15.72
15.72
16.18
16.40
16.67
17.00
17.55
17.83
18.00
18.16
18.18
18 . 34
18.3?
18.50
1E.65
19.00
19. 00
19.06
19. 2C
19.2C
19.23
15.30
19.50
19.50
15.56
19.66
19.70
19.70
19.80
20.00
2D.02
20.07
20.15
2.0.26
20.29
20.33
20.35
20.50
20.51
20.58
20.60
?0.67
20.96
20.96
20.96
20.96
20.96
21.20
21.33
21.45
21.52
21.58
21.61
21.66
21.67
21.70
22.QO
22.20
22.22
22.50
22.50
23.15
23.30
23.33
23.58
25.00
25.08
26.04
26.20
26.20
26.80
28.00
30.00
30.not
31.41
33.30
34.50
65.50$
165.06 f
243.66
140
-------
INTER-LABORATORY STUDY
POLLUTANT - S02
SAWPLE NUHBER - 2
* 112
TRUE-VALUE 64.40
MEAN 62.62
64.16
478
UNITS - MICROGRARS PER CUPIC METER
RANGE
VARIANCE
STD. DEV.
COEf. VAR
57.90
80.96
9.00
14.37
C.I.(UPPER) 64.29
C.I.(LOWE R) 60.95
SKEtfNESS -t.41
ACCURACY -.37
BATA IN ASCENDING ORDER
1.29 f
5. 009
7.000
10.700
17.360
25.43
27.400
28.33
32.150
35.10
35.400
39.30
44.77
45.81
47.000
47.77
48.03
49.50
52.40
53.00
53.70
54.50
55.00
55.02
55.02
55.50
56.40
57.40
57.50
57.64
57.64
57.64
57.64
57.70
59.50
59.55
59.66
59.90
60.14
60.24
60.26
60.26
60.26
60.26
60. £8
61.00
61.00
61.02
61.10
61.30
61.34
61.63
61.70
62.03
62.29
62.43
62.50
62.88
62.90
63.33
63.36
63.40
64.00
64.10
64.10
64.22
64.37
64.50
64.57
64.60
64.65
64.69
64.81
64.93
64.94
64.99
65.00
65.QO
65.oo
65.QO
65.50
65.50
65.50
65.50
66.22
66.43
66.53
66.67
66.67
66.83
67.QO
68.00
68.12
68.12
68.12
68.12
68.60
6E.69
68.77
68.80
69.00
69.30
69.30
69.44
70.00
70.00
70.20
70.70
70.80
71.00
71.88
72.92
73.36
74.40
75.00
75.20
75.20
75.92
76.80
78.74
83.33
132.700
230.560
579.020
141
-------
INTER-LABORATORY STUDY
POLLUTANT - S02
SAMPLE NUMBER - 3
* 111
TRUE-VALUE 64.70
*EAN 63.96
MEDIAN 65.00
DATA IN ASCENDING ODDER
478
UNITS - MICR06RAMS PER CUBIC METER
RANGE
VARIANCE
STD. DEV.
COEf.
73.00
117-00
10.82
16.91
C.I.(UPPER) 65.98
C.I.(LOWE R> 61.95
SKEWNESS -1.22
ACCURACY .46
1.29 •
5.30 f
17. 36-0
18.70 •
20.33
20.96
27.64 •
27.70
35.10
37. 60$
40.73§
45.81
47.16
48.03
48.84
48.92
49.00 •
52.40
53.20
54.90
55.02
55.02
57.50
57.64
58.00
58.06
58.42
59.00
59.50
59.70
60.00
60.15
60.24
60.26
60.26
60.50
61.00
61.57
61.67
61.C8
61.90
62.18
62.20
62.29
62.30
62.30
62.SO
62.as
62.90
63.00
63.00
63.20
63.36
63.70
64.00
fe4.10
64.10
64.22
64.57
fc4.5fc
64.75
64.99
64.99
65.OG
65.00
65.11
65.15
65.33
65.50
65.50
65.50
65.50
65. 50
65.5c
65. 5 C
65.52
65.77
65.88
66.Q5
66.32
66.33
66.43
66.67
66.89
67.40
67.43
67.58
68.00
68.12
68.12
68.52
68.76
69.00
69.30
69.44
70.QO
70.4o
70.58
70.70
70.74
70-83
71.00
71.70
72.00
72.50
73.12
73.20
73.36
73.36
75.20
76.36
76.80
77.70
78.54
78.60
78.62
91.00
91.66
93.33
225.32 •
710.02
796.70
w
142
-------
INTER-LABORATORY STUDY
POLLUTANT - S02
SAMPLE NUMBER - 4
1 111
TRUE-VALttE 126.20
WEAN 120.89
121.60
478
UNITS - HICR06RAHS PER CUBIC METER
RANGE
VARIANCE
STB. OEV.
COEf. VAX.
146.00
277-60
16.66
13.78
C.I.CUPPfcR) 123.99
C.I.(LOVE R* 117.79
SKEtiNESS .06
ACCURACY -3.65
»ATA IN ASCENBINS ORDER
3.17 f
10.90 •
52.03$
54.00
62.00 f
70.20
71.51
75.30 •
81.22
82.50
87.00 0
96. OOf
98.38 0
99.00
99.84
100.75 0
104. BO
105.00
106.60
106.70
107.20
107.42
107.97
109.41
110.00
111.33
112.50
112.66
112.67
112.82
113.50
114.13
115.28
116.00
116.67
116.84
116.85
117. 2D
117.59
117.80
117.90
117.90
117.90
118.00
118.00
118.40
118.64
118.75
118.80
119.00
119.00
119.02
119.09
119.64
119.98
12D.O*
120.20
120.50
120.52;
120.52
120.52
120.83
120.92
121.00
121.60
121.71
121.76
121.84
122.07
122.09
122.30
122.35
122.6?
123.02
123.14
123.14
123.14
123.14
123.14
123.31
123.8Q
124.QO
124.QO
124.10
124.96
125.QO
125.QO
125.76
125.76
125.76
125.87
126.80
126.80
127.60
128.30
128.38
128.46
128.69
128.97
129.16
129.42
129.49
129.60
129.83
129.98
130.21
130.66
130.70
130.76
133.20
133.60
136.00
138.86
139.64
141.48
145.92
149.22
165.00
165.06
200.00
343.22
1000.00
143
-------
INTER-LABORATORY STUDY
POLLUTANT - S02
SAMPLE NUMBER - 5
« til
TRUE-VALUE 204.00
1EAM 190.70
197.02
478
UNITS - HICR06RABS PER CUBIC METER
RANGE
VARIANCE
STD.
COEF
271.33
518-10
38.96
20.45
C.I
C.I.CLOWE
SKEUMESS
ACCURACY
197.95
R> 183.45
-1.89
-3.4?
BAT A IN ASCENDING ORDER
5.19
11.57
16.10
22.00
28.25
33.25
61.66
81.30
98.00
110.59
113.50
115.26
130.76
136.00
140.40
153.00
154.58
155.10
157.00
157.20
158.33
162.40
163.59
165.06
166.10
170.50
173.33
177.13
180.78
181.40
183.00
183.33
183.52
183.68
184.36
184.40
185.60
186.02
188.58
188.64
186.83
189.75
190.00
190.20
190.65
191.20
191.26
192.22
193.29
193.32
193.40
193.66
193.80
194.00
194*60
195.0C
195.40
195.6Z
196.50
196.5C
196.50
196.54)
197.00
197.02
197.02
197.50
197.56
198.00
19B.63
19B.7Z
199.00
199.12
199.80
200.00
2CO.QO
200.30
200.39
200.49
200.49
200.75
201.00
201.74
201.74
201.74
201.74
202.9Q
203.00
203.33
203.94
204.36
204.40
205.21
205.6Q
206.oo
207.47
208.QO
208.30
208.6o
209.32
209.60
2Q9.60
211.98
215.40
217.00
217.01
217.10
217.46
217.46
219.74
223.47
225.32
226.18
227.94
229.00
229.16
236.90
249.87
256.56
270.00
293.33
615.70
1000.00
144
-------
TECHNICAL REPORT DATA
(I'lcase read Instructions on ihe reverse before completing!
. REPORT NO.
EPA 600/4-80-017
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AMD SUBTITLE
SUMMARY OF AUDIT PERFORMANCE: MEASUREMENT OF S00,
N02, SULFATE, NITRATE, LEAD, HI-VOL FLOW RATE - T978
5. REPORT DATE
June 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Steven M. Bromberg, Robert L. Lampe, Berne I. Bennett
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Environmental Monitoring Systems Laboratory
Quality Assurance Division
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
A09A1D
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Monitoring Systems Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA 600/08
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report summarizes the results of the quality assurance audits for the
period from January, 1978 through December, 1978. Pollutants for which audits
were conducted and results reported are S02, N02, sulfate, nitrate, lead, and
hi-vol flow rate. The operation of the EPR audft program is also described.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI l-'icld/Group
Air pollution
Quality control
Ambient air
Criteria pollutants
68 A
43 F
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
144
20. SECURITY CLASS /Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |