EPA-600/4-77-043
October 1977
Environmental Monitoring Series
THE STATUS AND QUALITY OF RADIATION
MEASUREMENTS FOR AIR
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas. Nevada 89114
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U.S. Environ-
mental Protection Agency, have been grouped into five series. These five
broad categories were established to facilitate further development and
application of environmental technology. Elimination of traditional grouping
was consciously planned to foster technology transfer and a maximum
interface in related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
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
Information Service. Springfield, Virginia 22161.
-------
EPA-600/4-77-043
October 1977
THE STATUS AND QUALITY OF RADIATION MEASUREMENTS
FOR AIR
by
D. G. Easterly, R. R. Kinnison, A. N. Jarvis, and R. F. Smiecinski
Monitoring Systems Research and Development Division
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory-Las Vegas, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not consti-
tute endorsement or recommendation for use.
ii
-------
FOREWORD
Protection of the environment requires effective regulatory actions which
are based on sound technical and scientific information. This information
must include the quantitative description and linking of pollutant sources,
transport mechanisms, interaction, and resulting effects on man and his envi-
ronment. Because of the complexities involved, assessment of specific pollu-
tants in the environment requires a total systems approach which transcends
the media of air, water, and land. The Environmental Monitoring and Support
Laboratory-Las Vegas contributes to the formation and enhancement of a sound
integrated monitoring data base through multidisciplinary, multimedia programs
designed to:
• develop and optimize systems and strategies for moni-
toring pollutants and their impact on the environment
• demonstrate new monitoring systems and technologies by
applying them to fulfill special monitoring needs of
the Agency's operating programs
This summary report, "The Status and Quality of Radiation Measurements
for Air," should be useful in evaluating the quality of environmental radia-
tion data. The data contained in this report should be of value to the EPA,
other Federal agencies, State agencies, and private laboratories. For further
information on the data contained in this publication contact the Quality
Assurance Branch, Environmental Monitoring and Support Laboratory-Las Vegas,
Nevada.
Director
Environmental Monitoring and Support Laboratory
Las Vegas
iii
-------
ABSTRACT
As part of the radiation quality assurance program conducted by the
U.S. Environmental Protection Agency, calibrated radionuclide solutions are
distributed to participating laboratories for instrument calibration and yield
determinations. Laboratory performance studies involving the analysis of
radionuclides in environmental media are also conducted.
A summary is given of the results of the air filter cross-check program
for 1973-1975. Examination of these results indicate that gross alpha is the
least difficult C86 percent within the control limits for accuracy) and gross
beta is the most difficult (39 percent within the control limits for accuracy)
for the laboratories to analyze.
iv
-------
CONTENTS
Page
Foreword ill
Abstract iv
List of Figures and Tables vi
Introduction 1
Methods and Procedures 4
Preparation of Air Filters 4
Analysis by Participants 5
Results and Discussion 10
Summary 27
APPENDIX. Statistical Calculations 29
-------
LIST OF FIGURES
Number Page
1 Computer report 6
2 Control chart 8
3 Histogram of laboratory averages reported for strontium-90. . 15
4 Histogram of laboratory averages reported for cesium-137. . . 18
5 Histogram of laboratory averages reported for gross alpha . . 21
6 Histogram of laboratory averages reported for gross beta. . . 24
LIST OF TABLES
1 Summary of cross-check programs 3
2 Summary of air filter data for 1973-1975 12
3 Summary of laboratory performance,
1973-1975 laboratory intercomparison studies - air filters . 28
vi
-------
INTRODUCTION
Environmental radiation measurements are made routinely by Federal,
State, local, and private agencies. The data obtained from these measurements
are utilized by the U.S. Environmental Protection Agency (EPA) and other
agencies for such purposes as estimating dose and health effects, establishing
standards and guides, and conducting regulatory activities. Therefore, it is
essential that the precision and accuracy of the data be assured so that
policy decisions concerning environmental quality are based on valid and
comparable data.
The radiation quality assurance program of the EPA is designed to encour-
age the development and implementation of quality control procedures at all
levels of sample collection and analysis, and data processing and reporting.
As an integral part of the EPA's program, the Quality Assurance Branch of the
Environmental Monitoring and Support Laboratory-Las Vegas (EMSL-LV) distrib-
utes calibrated radionuclide solutions for instrument calibration and chemical
yield determinations, develops and tests analytical procedures for possible
use as EPA-approved laboratory methods, and conducts a number of laboratory
performance studies involving the analysis of radionuclides in environmental
media.
The Laboratory Intercomparison Studies Program (performance studies)
enables participating laboratories to maintain checks on their analyses and
assists them in documenting the validity of their data. In addition, this
program enables the EPA to obtain an overall estimate of the precision and
accuracy of environmental radiation measurements, and the precision and
accuracy of laboratory radioassay procedures for environmental samples.
Performance studies currently in progress involve samples of most envir-
onmental media and include milk, air, water, soil, diet, urine, and noble
gases. Table 1 is a summary of these programs. Participants include some
nuclear facilities and State, Federal and international laboratories. Because
of the large and growing number of participants and the continuing nature of
the programs, sufficient data are generated to permit periodic assessment of
the quality of environmental data.
Participating laboratories perform analyses on the cross-check samples
and return their data to the Quality Assurance Branch for statistical evalua-
tion. Comparisons are made between laboratories and within an individual
laboratory for accuracy and precision. A computer report and an updated
performance chart are returned to each participant. This enables each labora-
tory to document the "precision and accuracy of its radiation data, to identify
-------
instrumental and procedural problems, and to compare performance with other
laboratories.
A preliminary report on the laboratory performance studies conducted for
air filters is presented on the following pages.
-------
TABLE 1. SUMMARY OF CROSS-CHECK PROGRAMS*
Sample
Milk
Watur
Cross a, 6*
Gamma
'II
2"PU*
Kndium
Air
Gross o, 8*
*"PU*
Soil*
Diet
Urine
fins
89Sr,
l37Cs
Gross
60Co,
137Cs
3H
239Pu
22SRa
<*, 6,
239pu
238Pu
228Th
"Sr,
»37Cs
3H
8SKr
Analysis
90Sr, 131I,
, lu°Ba, K
o, B
106Ru, 13"Cs,
, 51Cr, 6SZn
, 2J8Ra
90Sr, 137Cs
, Z39Pu
, 230Th, 232Th
9«Sr, >»I,
, l"°Ba, K
Activity
Per Isotope
< 200 pCi/Hter
< 100 pCi/liter
< 500 pCi/liter
< 3500 pCi/liter
< 10 pCi/liter
< 20 pCi/liter
< 200 pCi/sample
< 2 pCi/sample
< 50 pCi/sample
c 200 pCi/kg
< 3500 pCi/liter
< 20 pCi/ml
Quantity
Supplied
"X. 4 liters
•v 4 liters
•x- 4 liters
% 50 ml
•x- 4 liters
•X" 4 liters
3 - 2"-diam.t
air filters
3 - 2"-diam.t
air filters
"x- 35 grams
2 - 4-liter
samples
•v 50 ml
10 liters
Preservative
Formalin
0.5JJ HN03
0.5N HN03
none
0.5N HN03
0.5$ HN03
none
none
none
Formalin
Formalin
none
Distribution
Bimonthly
Bimonthly
Bimonthly
Bimonthly
Semiannually
Quarterly
Quarterly
Quarterly
Semiannually
Quarterly *
Quarterly
Quarterly
Time for
Analysis
& Report
6 weeks
4 weeks
4 weeks
4 weeks
8 weeks
6 weeks
6 weeks
6 weeks
8 weeks
8 weeks
4 weeks
6 weeks
*l..jboratories are required to have the necessary licenses before receiving these samples.
tMi-trlc equivalent » 50.8 millimeters
-------
METHODS AND PROCEDURES
Glass fiber filters containing known amounts of specific radionuclides
are prepared and distributed to a number of Federal, State and private labora-
tories. These samples are designed to test the ability of participating
laboratories to analyze air filters for strontium-90, cesium-137, and gross
alpha and gross beta radioactivities. Eight air filter intercomparison stud-
ies were conducted during the period July 1973 - December 1975.
Three 2-inch diameter* (or 4-inch diameter upon request) simulated air
filters containing known amounts of radionuclides were sent to each partici-
pant. The filters were prepared to represent simulated air filter samples.
The radionuclides and the concentration ranges are as follows:
Isotope Concentration Range
90Sr 38-198 pCi/filter
137Cs 25-1180 pCi/filter
gross alpha 38-196 pCi/filter
gross beta 101-735 pCi/filter
PREPARATION OF AIR FILTERS
Two- and 4-inch diameter glass fiber filters are sprayed with six coats
of clear acrylic paint.t After the paint has dried, the filters are attached
to individual stainless steel planchets using short strips of double-sided
masking tape. The radioactive "spike" solution (after it has been checked for
accuracy) is pipetted, in the form of tiny droplets, uniformly onto the fil-
ters. The filters are allowed to air dry and then dried overnight in a 110° C
drying oven. After cooling to room temperature, the filters are analyzed for
radioactivity. If the calculated activities are within the 2a-range counting
statistics, the filters are packaged and shipped to the participating
laboratories.
*Metric equivalent = 50.8 millimeters
tKrylon #1301 Crystal Clear Acrylic Spray Coating
-------
ANALYSIS BY PARTICIPANTS
Participating laboratories conduct three independent determinations for
each radionuclide included in the particular cross-check sample and report the
results to the Quality Assurance Branch. Control limits previously estab-
lished by the Analytical Quality Control Service are used in analyzing the
quality of the results obtained by these laboratories. These limits are based
on the purpose for which the data are being obtained^ and on reasonable
laboratory ability. Upon receipt of the reports from all participating labo-
ratories, the data are analyzed using a computer. For each radionuclide, this
analysis includes determination of the following parameters: the experimental
average and standard deviation for each laboratory, the normalized range, the
normalized deviation from the known and grand average, and the experimental
sigma and the grand average of all laboratories. Examples of sample calcula-
tions to illustrate the computations performed by the computer are shown in
the Appendix.
A report is generated containing the results of the data analysis and the
data reported by all participating laboratories, listed according to labora-
tory identity code. Examples are shown in Figure 1. In addition, a control
chart is generated for each radionuclide included in the sample (Figure 2).
The control charts are updated each time a laboratory participates in a cross-
check study, thus giving each laboratory a continuous record of its perfor-
mance. A copy of the computer printout and a control chart for each radionu-
clide are mailed to each participant approximately 6 weeks following the
report due date.
-------
EMSLr-LV TRITIUM IN URINE CROSS-CHECK PROGRAM SEPTEMBER 1974
09/20/74 SAMPLE - A 3H
KNOWN VALUE = 3273 PCI/L
EXPECTED LABORATORY PRECISION (IS, 1 DETERMINATION) = 357 PCI/L
LAB RESULT
AN NO DATA PROVIDED
EXPERIMENTAL
SIGMA
CF
CF
CF
CM
CM
CM
CO
D
D
D
J
J
J
P
Z
Z
Z
3269
3522
3632
3261
3373
3362
NO DA:
3060
3060
3240
3255
3247
3294
NO DA1
3240
3340
3190
186.1
61.7
103.9
25.1
76.4
EXPERIMENTAL SIGMA (ALL LABS)
RNG ANLY
(R + SR)
0.60
0.19
0.30
0.08
0.25
149
AVERAGE
3474
3332
3120
3265
3257
NORMALIZED DEVIATION
(GRAND AVG) (KNOWN)
0.9
0.2
-0.8
-0.1
-0.2
1.0
0.3
-0.7
-0.0
-0.1
GRAND AVERAGE = 3290
Figure 1. Computer report
-------
Explanation of terms in Figure 1:
Title:
Column 1:
Column 2:
Column 3:
Column 4:
Column 5:
Column 6:
Column 7:
Bottom of
Chart:
Program name, sample collection date, sample code letter,
analysis type, known concentration of radionuclide,
expected standard deviation of analysis - single
determination.
Laboratory identification code (A, B, C, etc.).
Laboratory results (0-25 results listed down column).
One sigma (standard deviation) of the experimental
results.
Normalized range value in "mean range + standard error of
the range" (R + a ) units for comparability. (See
Statistical Techniques for Quality Control of Environ-
mental Radioassay, AQCS Report Stat-1, November 1964,
pages 4-8.)
(SR = a for printing purposes.)
K
Average value.
Normalized deviation from the grand average value of all
laboratories expressed in O units.
Normalized deviation from the known value expressed in O1
units.
M
One sigma experimental error of all laboratories, and the
grand average of all laboratories.
-------
<
10
9
8
7
6
5
4
3
2
1
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-10
R+3
B+2
B+1
B+0
TRITIUM IN URINE CROSSCHECK PROGRAM
LAB - D 3H
NORMALIZED DEVIATION FROM KNOWN
1973
1974
CL
WL
WL
CL
JFMAMJJASOND JFMAMJJASOND
NORMALIZED RANGE
1973 1974
I
CL
WL
JFMAMJJASOND JFMAMJJASOND
Figure 2. Control chart.
-------
Explanation of terms used in the control chart (Figure 2):
Title: Name of program, laboratory code letter, and type of
analysis.
Upper Graph: "Normalized deviation from known versus the month of
analysis." [ihe 95.0% (y,± 2aM) and the 99.7%
(u ± 3cr ) confidence levels were chosen as the warn-
JM. ~\
ing levels and control limits respectively.J
Lower Graph: "Normalized range values CR + OR) versus the month of
analysis." [ihe 97.5% (R + 2c£) and 0,100% (R + 3aR)
confidence levels were chosen as the warning levels
and control limits respectively .J
-------
RESULTS AND DISCUSSION
A laboratory is considered accurate, for our purposes, if its normalized
deviation (knoinO is within ±3. A laboratory is considered precise if its
range analysis (R + SR) is less than or equal to A*.
The results of the air filter intercomparison studies are listed in
Table 2 and shown in graph form in Figures 3-6. The x indicates the grand
average and the y denotes the known value. Each box represents an individual
laboratory result.
STRONTIUM-90 (FIGURE 3)
All eight studies contained a strontium-90 spike. Forty-five percent of
the laboratories reported results within the accuracy limits for the 3-year
period. Eighty-eight percent reported replicate results with acceptable
precision. In seven of the eight studies, the grand average value was less
than the known value. This may be an indication that the analysis is biased
low.
CESIUM-137 (FIGURE 4)
Cesium-137 was present in all eight studies. Fifty-five percent of the
laboratories reported data within the acceptable limits while 96 percent of
the laboratories possessed acceptable precision in their measurements. A
comparison of the known and grand average values does not show any bias in the
analysis.
GROSS ALPHA (FIGURE 5)
Seven of the eight studies included gross alpha activity. Eighty-six
percent of the laboratories reported results within the control limits, and
100 percent met the precision requirements. In five of the seven studies the
grand average values fell lower than the known values, possibly indicating a
negative bias in the analysis.
*Rosenstein, M., and A. S. Goldin, "Statistical Techniques for Quality
Control of Radioassay," AQCS Report Stat-1, U.S. Department of Health,
Education and Welfare, PHS, Nov 1964
10
-------
GROSS BETA (FIGURE 6)
Gross beta activity was present in seven of the eight studies. Thirty-
nine percent of the laboratories reported results within the control limits,
and 97 percent reported replicate results with a precision meeting the
requirements of the studies. In six of seven studies the grand average values
were greater than the known values which would imply that the analysis is
biased high.
11
-------
TABLE 2. SUMMARY OF AIR FILTER DATA
FOR 1973 - 1975
N
y(pd/f liter)
x(pCi/filter)
0(pCi/filter)
s(pd/f liter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
9
1180
1172
118
138
10
12
July 1973
9Q
Sr
5
90
81
9
14
10
16
Gross Alpha
Gross Beta
N
y(PCi/fliter)
x(pCi/filter)
a(pd/filter)
s(pd/f liter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
14
46
48
5
9
11
20
May 1974
90Sr
9
47
43
5
9
11
19
Gross Alpha
10
48
36
12
6
25
13
Gross Beta
14
194
178
10
19
5
10
June 1974
N
y(PCi/fliter)
x(pCi/filter)
a(pCi/filter)
s(pd/f liter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
18
200
199
10
31
5
16
90
Sr
8
198
187
10
16
5
8
Gross Alpha
15
196
177
49
39
25
20
Gross Beta
20
735
770
37
162
5
22
N = sample size; y = known value; x
and s = standard deviation.
grand average; a = expected precision;
(continued)
12
-------
TABLE 2. (continued)
October 1974
N
}j(pCi/f liter)
x(pCi/filter)
aCpCi/filter)
s(pCi/fliter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
24
174
184
9
18
5
10
9Q
Sr
15
172
164
9
18
5
10
Gross Alpha
29
* 174
171
44
65
25
37
Gross Beta
27
519
296
26
101
5
19
February 1975
N
]j(pCi/filter)
x(pCi/filter)
0(pCi/filter)
s(pd/f liter)
(a/y) x 100 (%)
(s/y) x 100(%)
137
Cs
20
151
161
8
19
5
13
90
Sr
13
160
164
8
24
5
15
Gross Alpha
25
150
144
37
42
25
28
Gross Beta
25
470
506
24
70
5
15
June 1975
N
y(pd/filter)
x(pCi/filter)
a(pd/filter)
s(pCi/fliter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
29
135
142
7
17
5
13
90
Sr
17
118
112
6
17
5
14
Gross Alpha
32
123
128
31
26
25
21
Gross Beta
27
371
389
19
44
5
12
N = sample size; y = known value; x = grand average; a
and s = standard deviation.
expected precision;
(continued)
13
-------
TABLE 2. (.continued)
N
]j(pCi/f liter)
x(pCi/filter)
or(pCi/filter)
s(pCi/filter)
(a/y) x 100(%)
(s/y) x 100(%)
137
Cs
27
99
96
5
11
5
11
September 1975
90Sr
11
98
91
5
9
5
9
Gross Alpha
36
99
89
25
22
25
22
Gross Beta
31
295
304
15
29
5
10
N
y(pCi/filter)
x(pCi/filter)
a(pCi/filter)
s(pCi/filter)
(0/y) x 100(%)
(s/y) x 1QO(%)
137
Cs
26
25
28
5
7
20
28
December 1975
90Sr
18
38
36
2
7
5
18
Gross Alpha
33
38
40
10
7
26
18
Gross Beta
26
101
107
5
8
5
8
N = sample size; y = known value; x
and s - standard deviation.
grand average; cr = expected precision;
14
-------
HH CONTROL
LIMIT
a OUTLIER
is, JULY 1973
FREQUENC
0 01 o
_ i i I
IT
l^H • n n s s
\ |_|J| 1 1
-------
o
Z 10-J
UJ
D
nj 5
C
u.
o
CONTROL
LIMIT
o OUTLIER
is-, OCTOBER 1974
X
H
I r i i
O 1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
is-, FEBRUARY 1975
1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
15-,
> 1 X
11°1 H
" 5 *
UJ & "
K
u.
O
JUNE 1975
i i i i i
O 1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
Figure 3 (continued). Histogram of Laboratory Averages Reported for
Strontium-90
16
-------
o
z
IU
D
UJ
CONTROL
LIMIT
a OUTLIER
SEPTEMBER 1975
x
\
r
n
1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
DECEMBER 1975
1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
Figure 3 (continued). Histogram of Laboratory Averages Reported for
Strontium-90
17
-------
CONTROL
LIMIT
a OUTLIER
15-
U
z 10.
Ul
D
u.
0.
.•P 1
JWk 1 IT/ O
^^
X
• : . • • — *... •.,-,_/ / .. ., , .n
9CX) 1OOO 11OO 12OO 13OO 14OO 155O
CONCENTRATION (pCi/filter)
U
a SH
K
u.
O'
MAY 1974
I
T
O 1OO 2OO 3OO 4OO 5OO 6CX)
CONCENTRATION (pCi/filter)
15- JUNE 1974
1OO £OO 3OO 4OO SCO 6OO
CONCENTRATION (pCi/filter)
Figure 4. Histogram of Laboratory Averages Reported for Cesium-137
18
-------
o
g 10^
D
uj 5 •
D
S 5^
o:
u.
O
CONTROL
LIMIT
a OUTLIER
OCTOBER 1974
n n
2OO 3OO 4OO 21OO 22OO
CONCENTRATION (pCi/filter)
FEBRUARY 1975
x
H
I
R
1 I i r
1OO 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
JUNE 1975
x
M
1
^^w^"
1(X) 2OO 3OO 4OO 5OO 6OO
CONCENTRATION (pCi/filter)
Figure 4 (continued). Histogram of Laboratory Averages Reported for
Cesium-137
19
-------
CONTROL
LIMIT
o OUTLIER
SEPTEMBER 1975
2OO 3OO 4OO 5OO
CONCENTRATION (pCi/filter)
6OO
DECEMBER 1975
n
1OO 2OO 3OO 4OO 5OO
CONCENTRATION (pCi/filter)
600
Figure 4 (continued).
Histogram of Laboratory Averages Reported for
Cesium-137
20
-------
u
z
OJ
D
01
o:
u.
CONTROL
LIMIT
D OUTLIER
MAY1974
100
i
200
300
i
40O
5OO
I
600
CONCENTRATION (pCi/filter)
JUNE 1974
100
200
3OO
4OO
6OO
7OO
CONCENTRATION (pCi/filter)
OCTOBER 1974
1OO 2OO 3OO 4OO
CONCENTRATION (pCi/filter)
i
5OO
600
Figure 5. Histogram of Laboratory Averages Reported for Gross Alpha
21
-------
15-1
15-1
CONTROL
LIMIT
a OUTLIER
FEBRUARY 1975
n
1OO
20O
30O
4OO
I
500
I
600
CONCENTRATION (pCi/filter)
JUNE 1975
3OO
1
4OO
500
I
600
CONCENTRATION (pCi/filter)
SEPTEMBER 1975
300
I
4OO
I
5OO
6OO
CONCENTRATION (pCi/filter)
Figure 5 (continued).
Histogram of Laboratory Averages Reported for
Gross Alpha
22
-------
CONTROL
LIMIT
D OUTLIER
DECEMBER 1975
i i i i I
1OO 2OO 3OO 4OO 500 6OO
CONCENTRATION (pCi/filter)
Figure 5 (continued). Histogram of Laboratory Averages Reported for
Gross Alpha
23
-------
>
1 1CH
2 5H
a:
u.
15-,
2 5H
a:
u.
CONTROL
LIMIT
a OUTLIER
MAY 1974
200
4OO
i
600
800
I
1OOO
1200
CONCENTRATION (pCi/filter)
JUNE 1974
FREQUENCY
10.
5 -
O .
n
I
O 2OO 4
6OO
8OO
1000
1600
CONCENTRATION (pCi/filter)
OCTOBER 1974
n n
2OO
4OO
6OO
800
1OOO
i
12OO
CONCENTRATION (pCi/filter)
Figure 6. Histogram of Laboratory Averages Reported for Gross Beta
24
-------
CONTROL
LIMIT
a OUTLIER
DECEMBER 1975
n
200
4OO
I
6OO
I
800
1OOO
1200
CONCENTRATION (pCi/filter)
Figure 6 (continued).
Histogram of Laboratory Averages Reported for
Gross Beta
25
-------
CONTROL
LIMIT
o OUTLIER
is-. FEBRUARY 1975
o x
ui 1°" p
| 5- '
U.
O . .
2OO 4OO 6OO 8OO 1OOO 12OO
CONCENTRATION (pCi/filter)
_ JUNE 1975
I I 9 T
6OO BOO 1OOO 12OO
CONCENTRATION (pCi/filter)
SEPTEMBER 1975
2OO 4OO 6OO BOO 1OOO 12OO
CONCENTRATION (pCi/filter)
Figure 6 (continued). Histogram of Laboratory Averages Reported for
Gross Beta
26
-------
SUMMARY
Table 3 is a summary of the results of the analysis for the air filter
cross-check samples. Using the percentage of laboratories reporting data
within the 3-sigma control limits as criteria, the radionuclides are listed in
the order of the ability of the laboratories to perform the radionuclide
analysis. The top chart refers to the ability of the laboratories to maintain
the required accuracy, while the bottom chart is a measure of the laborato-
ries' ability to meet the precision requirements.
The conclusions drawn, of necessity, have been very general due to the
limited amount of available data. The data indicate that gross alpha is the
least difficult (86 percent within the control limits for accuracy and
100 percent within the control limits for precision over the 3-year period)
for the laboratories to analyze. Gross beta is the most difficult to analyze
(39 percent within the control limits for accuracy and 88 percent within the
control limits for precision).
With the continuation of these studies, additional data will be collected
and compiled. When more data become available, such parameters as control
limits, methods of analysis, and instrument calibration will be critically
assessed.
27
-------
TABLE 3. SUMMARY OF LABORATORY PERFORMANCE
1973-75 LABORATORY INTERCOM?ARTSON STUDIES -
AIR FILTERS
% of Laboratories within ± 3(J
Radionuclide Analysis C99.7% Control Limits)
1973 1974 1975 1973-75
Gross Alpha 82 88 86
Cesium-137 90 53 53 55
Strontium-90 44 49 42 45
Gross Beta 32 42 39
% of Laboratories Meeting the Expected
Radionuclide Analysis Laboratory Precision Requirements
1973 1974 1975 1973-75
Gross Alpha 100 100 100
Gross Beta 96 98 97
Cesium-137 100 95 96 96.
Strontium-90 86 93 85 88
28
-------
APPENDIX
STATISTICAL CALCULATIONS
To illustrate the computations performed by the computer, example calcu-
lations are given using data for three samples analyzed at one laboratory
(Laboratory D, see Figure 2).
The experimental data are listed and the mean, experimental sigma, and
range are computed. These statistics provide measures of the central tendency
and dispersion of the data.
The normalized range is computed by first finding the mean range, R, the
control limit, CL, and the standard error of the range, 0. The normalized
range measures the dispersion of the data (precision) in such a form that con-
trol charts may be used. Control charts allow one to readily compare past
analytical performance with present performance. In the example, the normal-
ized range equals 0.3 which is less than 3, which is the upper warning level.
The precision of the results is acceptable.
The normalized deviation is calculated by computing the deviation and the
standard error of the mean, a . The normalized deviation allows one to readily
measure central tendency (accuracy) through the use of control charts. Trends
in analytical accuracy can be determined in this manner. For this example,
the normalized deviation is -0.7 which falls between +2 and -2, which are the
upper and lower warning levels. The accuracy of the data is acceptable.
Finally, the experimental error of all laboratories, the grand average,
and the normalized deviation from the grand average are calculated in order to
ascertain the performance of all the laboratories as a group. Any bias in
methodology or instrumentation may be found from these results.
29
-------
EXAMPLE CALCULATIONS (Laboratory D Data, see Figure 2)
Experimental data:
Known value = u = 3273 pCi 3H/liter urine on September 24, 1974
Expected laboratory precision = 0 = 357 pCi/liter
Laboratory Sample Result
D
D
D
xi
x2
x3
3060 pCi/ liter
3060 pCi/ liter
3240 pCi/liter
Mean = x
N
E -
x » -i^L -
where N = number of results = 3
Experimental sigma == s
= 3120 pCi/liter
N-l
/(3060)2 + (3060)2 +
•V
(3240)2 - C3060 + 3060 + 3240)
= 103.9 pCi/liter
Range = r
r = I maximum result - minimum result
3240 - 3060 = 180 pCi/liter
30
-------
Range Analysis (RNG ANLY)*
Mean range = R
R = d2a where d2* = 1.693 for N = 3
= (1.693)(357)
= 604.4 pCi/liter
Control limit = CL
CL = R + So,
K
= D^R where D4* = 2.575 for N = 3
= (2.575)(604.4)
= 1556 pCi/liter
Standard error of the range = aD
K
0R = (R + 3aR - R) v 3
= (Djj - R) -r 3
= (1556 - 604.4) -r 3
= 317.2 pCi/liter
Let range = r = wR + XCL = 180 pCi/liter
K
Define normalized range = w + x
for r > R, w = 1
then r = wR + xaR = R + xaR
r - R
or x =
*Rosenstein, M., and A. S. Goldin, "Statistical Techniques for Quality Control
of Environmental Radioassay," AQCS Report Stat-1, U.S. Department of Health,
Education and Welfare, PHS, Nov 1964
31
-------
r — R
therefore w+x= l+x= 1+
CTR
for r < R, x = 0
then r = wR + xa_, = wR
K
r
or w =
R
therefore w+x=w+0= ——
R
since r < R (180 < 604.4)
180
w + x =
604.4
= 0.30
Normalized deviation of the mean from the known value = ND
Deviation of mean from the known value = D
D = x - y
= 3120 - 3273
= -153 pCi/liter
Standard error of the mean = O
m
a --S-
m
N
357
= 206.1 pCi/liter
32
-------
am
-153
206.1
-0.7
Experimental sigma (all laboratories) = s CSee Figure 2)
N
(xj2 -
i N
N-l
/
162639133C49345)'
15
14
- 149 pCi/liter
Grand average = GA
N
N
49345
15
- 3290 pCi/liter
Normalized deviation from the grand average = ND1
Deviation of the mean from the grand average - D1
D' = x - GA
= 3120 - 3290
= -170 pCi/liter
33
-------
»•-
m
-170
206.1
= -0.8
34 «U.S. GOVERNMENT PRINTING OFFICE: 1977-786-181
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-77-043
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
THE STATUS AND QUALITY OF RADIATION MEASUREMENTS
FOR AIR
5. REPORT DATE
October 1977
6. PERFORMING ORGANIZATION CODE
y.AUTHORis) D> G_ Easterly, R. R. Kinnison,
A. N. Jarvis, and R. F. Smiecinski
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Las Vegas, Nevada 89114
10. PROGRAM ELEMENT NO.
1HD621
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency—Las Vegas, NV
Office of Research and Development
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
13. TYPE OF REPORT AND PERIOD COVERED
Interim, 1973-1975
14. SPONSORING AGENCY CODE
EPA-ORD-Office of Monitoring
and Technical Support
15. SUPPLEMENTARY NOTES
16. ABSTRACT
As part of the radiation quality assurance program conducted by the
U.S. Environmental Protection Agency, calibrated radionuclide solutions are
distributed to participating laboratories for instrument calibration and
yield determinations. Laboratory performance studies involving the analysis
of radionuclides in environmental media are also conducted.
A summary is given of the results of the air filter cross-check program
for 1973-1975. Examination of these results indicate that gross alpha is the
least difficult (86 percent within the control limits for accuracy) and gross
beta is the most difficult (39 percent within the control limits for 'accuracy)
for the laboratories to analyze.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
radiation chemistry
quality assurance
quality control
water analysis
isotopes
radium-226
tritium
chromium-51
cobalt-60
zinc-65
ruthenium-106
cesium-134
cesium-137
laboratory performance
intercomparison studies
cross-check
gross alpha
gross beta
07E
12B
14D
18B,D,H
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
44
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
------- |