f/EPA
United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
P.O. Box 93478
Las Vegas NV 89193-3478
EPA/600/R-93/005
January 1993
Research and Development
Uranium in Water
Intercomparison Study
A Statistical Evaluation of the
November 13,1992 Data
-------
Uranium in Water
Intercomparison Study
November 13, 1992
Environmental Protection Agency
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY-LAS VEGAS
P.O. BOX 93478
LAS VEGAS, NEVADA 89193-3478
(702/798-2100)
Dear Participant,
Enclosed are the results of the Nuclear Radiation Assessment Division (EMSL-LV)
Intercomparison Study for Uranium in Water; November 13, 1992.
The known value for each analysis was determined by gravimetric methods,
checked by chemical analyses performed by EMSL-LV's Radiochemistry Laboratory,
and compared to the participating laboratories' grand average.
The expected precision, determined by the known value, was taken from 'Table
3. Laboratory Precision: One Standard Deviation Values and Control Limits for
Various Analyses", which is based on data accumulated over the years by the
Intercomparison Program, and can be found in the Environmental Radioactivity
Laboratory Intercomparison Studies Program and Radioactive Standards Distribution
Program information brochure.
Please take a few minutes to review this report and the analytical data your
laboratory submitted to us. If there are any apparent discrepancies, please notify
us immediately.
We encourage you to make use of the computer-automated data-entry system that
has been in place for some time now. As the number of participants increases, and
it becomes unrealistic for us to receive results by mail or FAX, the computer system
will be our only avenue for accepting data.
If you have any questions or comments, please send a message via the data-entry
system or contact Frank Novielli at 702/798-2159 or Patricia Honsa at 702/798-2141.
Sincerely,
Frank Novielli
Senior Chemist
Radioanalysis Branch
Enclosure
-------
NOTICE
This material has been funded wholly by
the U.S. Environmental Protection Agency.
It has been subject to the Agency's review,
and has been approved for publication as
an EPA document.
-------
EMSL-LV Intel-comparison Study: Uranium in Water, 13-Nov-1992 1 / 7
The following pages consist of separate sections for each of the nuclides in this study with four
parts per section. After the first, each part is separated from the next by a new page or a thick
horizontal bar. The first page of each section is a statistical summary for the nuclide and starts
with a statement of the known value, the control limits, and the warning limits.
The warning limits are placed at two normalized standard deviations above and below the
known value and the control limits are three normalized standard deviations above and below
the known value. If you keep control charts, these values will be useful for anticipating problems
with the accuracy of your analytical methods.
The coin shaped pie chart at the top of the summary page shows the fate of all the samples
sent out in number and percentage terms. The pie chart starts at the top and rotates clockwise.
The first sector represents those participants who submitted analytical results within both the
warning and control limits. The next sector represents those who are in the warning region but
not out of control. The third sector represents those who are out of control, but have passed
the outlier test. The fourth sector represents those who have failed the outlier test. The last
sector represents those participants who have failed to respond properly. This is the case if no
analytical results were returned, or less than three determinations were reported, or if the results
were received too late. The reeding on the edge of the coin is spaced at one percent intervals, and
the sector shading becomes darker as the data reliability decreases. Sectors with zero width are
not shown.
The table in the center shows a number of statistical quantities calculated from the submitted
data based on the mean and median values in relation to the known value, both before and after
outlier removal. The lower pie chart uses the same construction as the upper chart and shows
the distribution of properly submitted data in terms of deviation from the known value divided
into sectors representing one, two, three, and greater than three normalized standard deviations.
The second part is an alphabetical listing, in lab-code order, of submitted data and several
calculated quantities. An entry that is shaded has been rejected because of one of the reasons
listed above or failure of the outlier test. The fifth and sixth columns are a measure of laboratory
precision. The Range 'analysis is a normalized value that you may use to keep precision control
charts. If this value is between 2.0 and 3.0, your analytical process precision is in the warning
zone;.if it exceeds 3.0 it is out of control. The eighth and ninth columns are the differences from
the mean of all non-outliers and from the known value, respectively. A tag symbol may appear in
the last column. Each page with tags has a symbol definition summary at the bottom. If there is
no tag symbol, the data is within the control limits, but it may be in the warning zone.
The third part is a three-column listing of result average, tag symbol, and lab-code in average
order excluding those labs not responding properly. In this order, all outliers and out-of-control
retmlts appear at the top or bottom of the list.
The last part is two bar chart displays showing frequency distributions of responding
participants. The first chart places the known value at the center and a bar at each 0.2 unit of
expected precision. The second chart places the mean of the reported measurements at the center
and a bar at each 0.2 unit of standard deviation. In both cases, a bar includes those results within
0.1 unit up to the maximum of six. Any results more than six units from the center value are
shown cumulatively by a shaded bar one past the sixth unit. If the central tendency of the known
value distribution falls away from the center, an error in accuracy is indicated. If the distribution
is broad, poor precision is indicated. The mean value distribution is similar but uses the average
and standard deviation of reported results as its basis.
-------
2/7 EMSL-LVIntel-comparison Study: Uranium in Water, 13-Nov-1992
The Range Analysis(R + SR) is calculated from the range, mean range and standard error of
the range values. The range is the difference between the maximum and minimum results for the
laboratory. The mean range is calculated by multiplying the expected precision by 1.693(for three
results). The standard error of the range is calculated by multiplying the mean range by 2.575(for
three results), subtracting the mean range from this product, and dividing the result by 3. If the
range is greater than the mean range, then the range analysis is calculated by subtracting the
mean range from the range, dividing the result by the standard error of the range and adding 1.
If the mean range is greater than or equal to the range, then the range analysis is calculated by
dividing the range by the mean range.
The normalized deviation of the mean from the grand average is calculated from the deviation
of the mean from the grand average and the standard error of the mean values. The deviation of
the mean from the grand average is calculated by subtracting the grand average from the
average of the laboratory's three results. The standard error of the mean is calculated by
dividing the expected precision by the square root of 3(the number of results). The normalized
deviation of the mean from the grand average is calculated by dividing the deviation of the mean
from the grand average by the standard error of the mean.
The normalized deviation of the mean from the known value is calculated from the deviation of
the mean from the known value and the standard error of the mean values. The deviation of the
mean from the known value is calculated by subtracting the known value from the average of the
laboratory's three results. The standard error of the mean is calculated by dividing the expected'
precision by the square root of 3(the number of results). The normalized deviation of the mean
from the known value is calculated by dividing the deviation of the mean from the known value by
the standard error of the mean.
A complete explanation of the statistical calculations involved in the report may be found in the
Environmental Radioactivity Laboratory Intel-comparison Studies Program information brochure
[Draft Revision of EPA-600/4-81-004], available from Frank Novielli, EMSL-LV, 702/798-2159.
-------
EMSL-LV Intercomparison Study: Uranium in Water, 13-Nov-1992
3/7
Uranium (Natural) Statistical Summary 101 Participants
The known value of this nuclide is 15.2 pCi/I with an expected precision of 3.0; the control limits
are 10.0 to 20.4; the warning regions are 10.0 to 11.7 and 18.7 to 20.4
27(26.7 %) Failed to respond
7(6.9 %) Outliers -'
1(1.O %) Out of control »
but not an outlier
Statistic
, 61 (60.4 %) Within all limits
Respondents
5(5.O %) In warning zone
but within control
Non-outliers
Mean
Std. Dev.
Variance
%Coef. ofVar.
% deviation of mean from known value
Norm. dev. of mean from known value
Median
% deviation of median from known value
Norm. dev. of median from known value
15.40
7.29
53.08
47.31
1.31
0.03
14.68
-3.40
-0.07
Grand Avg 14.44
1.74
3.04
12.07
-5.02
-0.44
14.67
-3.51
-0.31
8(10.8 %) More than 3 norm. S.D.
5(6.8 %) Between 2 and 3 norm. S.D. *
47(63.5 %) Within 1 norm. S.D.
of known value
14(18.9 %) Between 1 and 2 norm. S.D.
-------
4/7 EMSL-LV Intercomparison Study:
Uranium (Natural)
Lab Res. 1 Res. 2
AA
AF
AK
AL
AW
BA
BC
BK
BL
BO
C
CA
CE
CJ
CK
D
DE
DO
DT
DZ
E
EB
ER
FE
FN
FZ
GN
GZ
HK
HL
15.7
14.7
17.4
13.4
7.0
14.7
14.5
i; ^c,25;- '- '" '-->-
13.4
12.8
16.1
14.8
16.0
15.1
14.8
16.2
-^10^
17.3
14.4
16.4
13.0
15.0
16.0
14.0
15.6
15.5
14.6
13.4
,_14.8;,,
12.6
13.3
17.1
15.6
13.6
17.6
15.0
11.0
15.9
16.0
15.0
//?y
-------
EMSL-LV Intercomparison Study:
Uranium (Natural)
Lab Res. 1 Res. 2
MS 11.0
- jp^^|v;^:--^;^r;; - -
NA 15.0
NK 9.7
NO 13.8
NT 15.9
s"!^' ^'rrv^MT'tv^ -% """" -X?".
11.1
;
-,^\x ,;-
16.0
9.9
14.2
15.7
£,!-- ^r.
;5^(ili^i/^«^^4;l^i^
OM* ^WC^W; ?s^>
\ ^sv. '"'i S'v' v.- "" V" \-.« \>..v. -.V.V.. ".%' ...'% >l*s
OS 9.6
OY 16.0
OZ 16.6
PB 15.8
I^SR'SP&kS
>rapt:;^Sjt::^
;p£Hj.:Vv
A^v-fc-
""""""lb.3
15.6
16.4
15.2
V * 'O «.VH "> * i *ff\&- s
\s , S*i^ \^ - ',"
% *\^ ^V v " - s
**i!-SV.Si* ^ « -. * *"
^-VVvf^-1^ ::-\' x;;
- ^-V-t;s? ,^kV^ *~i.*^^*-"
,,"'--," ;--'^>
15.1
11.7
^88.9;'
17.0
5^, <-v;
12.7
10.2
i:::^x>
rtv^y',- : ^>.;
16.0
12.9
~':>T?,^1 A-^~
,\-\>;; -,v .* *> " -\vs
16.9
; vrv^-'s.r^vsv"-'
-';:; :,f'"' \s;;" '-
12.7
10.6
^ - *'<-w"8S!M^N *jsst*s*\v ^^ ^%v *
; v--^ v- , -. v.^'.-ixv-.V * ^v
Exper.
Uranium in Water, 13-Nov-1992
Rnganal
5/7
Normalized deviation
Sigma (R + SR) Average
0.06
1.00
0.20
0.45
0.20
V. , ,,>«^VJ.^. ^, >
,/^^v
! ^^\ ^;
0.40
0.26
0.20
0.30
v^C- \V<
""N** ""
0.47
0.92
.TTi;58" \
0.21
* \s ^ i
0.06
0.45
-- s s - vs^^s s ^-
0.020
, ^^"^vr x' "\
0.394
0.079
0.177
0.079
- y*»^-"< ,--- -^;'--_
^"v\'- ^ "" \S X *'''
°,;/;^^5ScV₯-
0.138
0.098
0.079
0.118
- ' V\i'-!x,"\x
'>' v - "", ,\,, A,.'
- \ \x-;>- \~
0.177
0.354
^6jB82--r:;
,.0.079
f ><v c s- ^ "*sy
0.020
0.177 .
, --"^r^^r
: ' QvAl^S?. !> ' ->-;C-^ \-"3;C ">'- vt ' ' ' V, VSS """ v ;i» 5^' ;;S^ :>
1 K6'W;XS'^.^"^V^s?>v4
^R^^SlK^aii'^fJ^
RM 13.5
RN 11.9
HP 15.8
RR 14.6
x *r vr^s'^^t^^^^A"
"s ^ ^N\-- ^ * ^ v tft , * *"isV
RW 13.9
RZ 14.7
_J_v__13.5
:^SCS^t'4lr-x--^§^[
X8D^_ ^12.7^7 ^
I fiLlf"^ y^y^f ^p
SOT '"""14.5" "'"
SS 14.4
;i»^:^ppgii
SZ 16.4
''wistr?;!;^!^8^^^
* TAi^ '^js^ s^Kt>%> ^'.
t ^l*s^ V* ; W^fe^
r--m^;^#;^lt:^
-TB^^^SltC^^
13.5
%N s "\ %
*s>^.-"-
' -i1' ''"-'V
9.9
13.3
15.8
14.7
^xSM
O * "-X * v%y
13.6
14.2
13.4
12.9
|§Silt
14.4
14.5
HW;
15.0
f^^WS
SVt*.">1*
14.2
"- "'>£ ^ -,--^
^v^:^;-, ,>^
J:^;2il^>
10.0
12.9
15.8
..WjR.. ,
^\ ^^'^^''T'^4'^!'' '"'"'' ^s
TS ^ "i"^ * " ' s .s^ i'* *% s'»
13.7
15.1
15.9
^^^S5r
\ ^7*12.9
^^^5 -W' A-U:
14.4
14.2
0.47
S SS S*M "
\V"
* V
;^6.34/-
2.05
0.72
0.00
0.10
"1 ' ">T~
0.15
0.45
_1.42 _^
?HTV^ r«"'\
0.12
1^ - ^"~ '' V
0.06
0.15
0.177
Ss s 'V
-.* s "* l " ,
\ - ^ ^-" . % ^-^
.V.8,Wx-'---*
0.709
0.276
0.000
0.039
^ %t s^" ^ s %., °
s «. %\ s ^J%
0.059
0.177
,^0.492 __^
^r^^^j
0.039
, sV^xv" * -~:
0.020
0.059
11.03
'''sV ' Vi
16.00
9.70
14.23
15.70
\",V - . " *\m
(grand-avg) (known) Tag
-1.97
- ^5jr^^>x>
0.90
-2.74
-0.12
0.73
-^\^V ^^"s" ^T ^ **° --'Xv
! 5 " i^-Jx^^i.^-V^^
:x;yv,c-"""!~v-,;
- ^ ^ ^ ^ - \^
10.07
15.70
16.60
15.50
^^\X;,
s\* ->«v\ \s
~- ; / >s- * '
15.47
11.90
760;83»':
17.07
^~-\. -;; .
12.73
10.63
<","!M& ,' f" '
- "V" '-"
13.67
^ N '
28^27
11.13
12.70
15.80
14.60
*f'-s^-' ~"-f
13.73
14.67
JL4.27_
\s" *;<%\^
12.83
xv, 1? ' '" Vy
14.43
14.37
- -v*^ '7,-',^!'. ^V
-2.52
0.73
1.25
0.61
- V *-'.'. .''".;
^ ^ S s " ^ ' "" s - V -'
J;j4>"iV 1
O s \
0.59
-1.46
.:? 31:98 ;^'.*;x
_^ 1.52^ ^
s\" " ""' ^ "X-""T^
-0.98
-2.20
" . * ^-.rv ^' -% "". s -
, # o - % - -\
-0.44
' \ V
^:"V.9fr '« "
-1.91
-1.00
0.79
0.09
% ''' ,
% s "" l
-0.41
0.13
,,,.r:5L10.». .
XV^;'"'- "'A" " ^** ^ \
-0.93
^-"'V H""'%v/-
0.00
-0.04
PUf^1^, ^"/';---:-; ^<4F?";.-yu^c/f:>^v
14.5
!&%%*°< Js^«&^;>;
-s-c->N!* .\ ,₯-? % :
^>^ti^^^^^ S' A^
^v O,1- X?^
""**>? ^k ^
:30i7f
«^V>?i>^«
lf%l$il';;f;
s No data submitted
0 s Insufficient data
0.98
0.374
»" * N 0 \, *^.X * "V" --.- v^O* J N 'V**S --f
.,>- " ^ t^ X<> t s % ^ % ^ w ? v> s V, *%.
^ * 'O% s%
x\^»*;->! *<,,V'x
"*<** * > ^\^' *.
- -txV\>'^> ." "s %
"^-^^-.^tfT': - /-^ ---;
^V1 " " NA s/ v.%
*>**&"*
;--o:847\.^".
15.30
"-7",.'^ ,»
" -;.S,-?x'';>"
A\; , , ^wr'
;28.87 ;
TAG SYMBOLS
x 3 Determined
to be an outlier
0.50
;- ^ ~_ ,s < -,«* -, -
'-a,^«,'x-< "!-\ \ *is"»>
,-2.41^^ ^
:^0 ''\vV ' ;
0.46
-3.18 li
-0.56
0.29
r ---^ s - - - ;
*/ ^ ^ ^ ^ ^
^ " V * s s ""
>cf?C--*-
-2.96
0.29
0.81
0.17
\ \ .-'" '
^ --> N , '
'
0.15
-1.91
3164 x
1.08
-1.42
-2.64
> -
s%
-0.89
' X
'7.54xo x.
-2.35
-1.44
0.35
-0.35
' .' ''' : '
-0.85
-0.31
-0.54
'V->;^--\- S; "
-1.37
, \ s '
-0.44
-0.48
t ',t ',, ,i \ ' «
0.06
V-s'V ", -
-\ :*>,> -
-''-- ~-.v'£ \\--~ ~>-:~
"T!"'8;33;--v
It 3 Above
11 = Below
- ~\,
7.89 x
control limit
control limit
-------
6/7 EMSL-LV Intercomparison Study: Uranium
Uranium (Natural)
Lab Res. 1 Res. 2
-.TGx'.- ";;8»7r*"" >6i7
TI 14.4 13.4
TJ ^:\Y'" y~' '; y
m* -,/ -\ H
s , V
TL 17.4 16.2
TP 14.9 14.7
TU 18:0_ 18.6
'^'i^-ViVV^f'c"''
Average Tag
6.10 x
7.10 x
7.50 x
9.70 U
10.07
10.63
11.00
11.03
11.13
11.90
12.00
12.70
12.73
12.80
12.83
13.07
13.10
13.30
13.33
13.43
13.67
13.73
13.93
14.00
3 No data submitted
0 s Insufficient data
Exper. Rng anal
Res. 3 Sigma (R + SR)
8.9 0;69 0.238
14.9 0.76 0.295
',-. " " " "* % V--
15.8 0.83 0.315
14.9 0.12 0.039
18.7 0.38 0.138
.^"l.'-.^'/c"'".. - : Y - '
in Water, 1 3 -No v- 1992
Normalized deviation
Average (grand-avg) (known) Tag
6; 10 \ - - >4i8i '- *' ;: -5u26>
14.23 -0.12 -0.56
". ;''.' - ' ' "~~
16.47 1.17 0.73
14.83 0.23 -0.21
18.43 2.31 1.87
*^
X * *'
;>.
*
*
Data sorted by Laboratory Average
Lab Average Tag
TG 14.00
DT 14.20
AW 14.23
NK 14.23
OS 14.23
QQ 14.27
LH 14.30
MS 14.37
RM 14.43
PV 14.53
ER 14.60
RN 14.67
QJ 14.67
BO 14.70
SD 14.70
AL 14.73
C 14.83
HY 15.13
IE 15.13
EB 15.30
QZ 15.33
RW 15.37
E 15.43
JY 15.47
15.47
TAG SYMBOLS
Lab Average Tag
FZ 15.47
GZ 15.50
TI 15.57
NO 15.67
BA 15.70
S 15.70
BC 15.77
SS 15.77
SO 15.80
ME 15.90
RR 15.97
RZ 15.97
FE 16.00
CK 16.10
CE 16.10
D 16.27
TP 16.47
L 16.60
HK 17.07
SZ 17.07
CJ 18.43
HP 24.60 x
AA 28.27 x
PP 28.87 x
JS 69.83 x
ft s Above control
x s Determined to be an outlier JJ a Below control
Lab
CA
PB
GN
FN
NT
OY
LZ
HL
RP
LT
AK
AF
NA
K
DE
J
TL
OZ
PX
DZ
TU
BK
RK
TE
PW
limit
limit
-------
EM8L-LVIntel-comparison Study: Uranium in Water, 13-Nov-1992 7/7
Uranium (Natural)
% responding labs Frequency distribution
10-
in
-2 d +2 +4
Normalized deviation of the mean from the known value
% responding labs Frequency distribution
10
1..
': -i " mill
1 -2 ' 6 r
6 -4 -20+2+4 +6
Normalized deviation of the mean from the grand average
------- |