- Email: [email protected]
Metrological capability of the Brazilian laboratories of analyses of radionuclides in environmental samples
Applied Radiation and Isotopes 56 (2002) 409–414
Metrological capability of the Brazilian laboratories of analyses of radionuclides in environmental samples L. Tauhata*, M.E.C. Vianna, A.E. de Oliveira, A.C.M. Ferreira, C.C.S. da Conceic-a* o ! Laboratorio Nacional de Metrologia das Radiac-oes * Ionizantes (LNMRI), Instituto de Radioprotec-ao * e Dosimetria (IRD), Comissao * Nacional de Energia Nuclear (CNEN) Av. Salvador Allende, s/n Recreio dos Bandeirantes - Rio de Janeiro C.E.P. 22780-160, Brazil Accepted 30 July 2001
Abstract The metrological capability of 24 Brazilian laboratories of radionuclide determinations in environmental samples was analysed for purposes to establish requirements for accreditation of laboratory for each radionuclide. For each type of analysis, the range of activity per unit mass or volume, the range of the reference value and its correspondent uncertainty were described. The accuracy was established using the statistical criteria of USE.P.A. for the laboratory performance classification. The precision of the analyses, expressed by its variation coefficient of the laboratory determinations, was classified in intervals of 5%. The data base was constituted by 3013 results of analyses obtained in 28 intercomparison runs performed from 1991 to 2000, for K, Th, Unat, gross alpha, gross beta, and other 27 radionuclides, in matrices of water, soil, vegetation, air filter and milk. The quality criteria established were the variation coefficient less than 5% and good performance in intercomparison runs. Results included from 5% to 10% are also considered, for special types of matrices and radionuclides. r 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction The increasing of environmental samples analyses due to the exigency of quality control programs and studies of the environmental impact in many types of the human activities, claim to a fast response of the laboratories and the quality of the results assured by normative or accreditation organisms. The laboratories need to present results with accuracy and precision met recommendations of this type of institutions and, moreover, to submit its technical procedures for accreditation process. In the case of the laboratory analyses of radionuclide, it is important to note the necessity of an accreditation process for each kind of radionuclide. Large efforts are made in the past time to develop a reasonable quality *Corresponding author. Tel.: +55-2134-118185; fax: +552144-21605. E-mail address: [email protected] (L. Tauhata).
criteria to be applied in the classification of the laboratories and to evaluate its measurement performance. In the case of the laboratories that analyzing environmental samples the repeatability of measurements and their normalized standard deviation are seldom-included in performance evaluation, which should be included to study the long term stability of a laboratory performance. Then, one good criteria that can be obtained is to select the group that has the best performance over long period or many intercomparison runs. Another important requirement is related to the precision of the measurements, expressed by the variation coefficient. The low value of variation coefficient is, obviously, related to the low uncertainties of the results. By combining these two criteria, it is possible to establish the requirement to select the laboratories with confiability of measurement, for each radionuclide. Besides, it is important to take account of the main objective of the
0969-8043/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 9 - 8 0 4 3 ( 0 1 ) 0 0 2 2 3 - 8
410
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
measurement, in respect of the precision need. Other exigencies, such as the difficulties of doing the analysis, the type of sample matrix, type of radiation of emitter, and measurement procedure should be also considered. Another important point is to assure the traceability of the laboratories measurements to the national and international metrology network and to the BIPM.
2. Methodology To evaluate the performance in terms of accuracy of 24 Brazilian laboratories, the statistical criteria of Environmental Protection Agency of United States, USE.P. A, was used for the laboratory performance classification (Jarvis and Siu, 1981). For each type of 100
100
80
Ac-228 : 24 Am-241: 25 Ba-133: 162 Bi-212: 18
60 40
Frequency (%)
Frequency (%)
80
0
0
0
5
10 15 20 Variation Coefficient (%)
Ac-228
Am-241
25
Ba-133
0
30
Ce-144
Co-57
20
25
Co-60
100 80 Cs-134: 187 Cs-137: 316 Gross alpha: 195 Gross beta: 225
60 40
Frequency (%)
Frequency (%)
10 15 Variation Coefficient (%)
Bi-214
80
20
H-3: 99 K-40: 71 Mn-54: 15 Pb-210: 97
60 40 20
0
0 0
5
Cs-134
10 15 Variation Coefficient (%)
Cs-137
20
Gross alpha
25
0
5
Gross beta
10 15 Variation Coefficient (%)
H-3
K-40
Mn-54
20
25
Pb-210
100
80
Pb-212: Pb-214: Pu-238: Pu-239:
60 40
Frequency (%)
100
Frequency (%
5
Bi-212
100
21 23 6 27
80 Ra-226: 186 Ra-228: 139 Ru-106: 166 Sb-125: 11
60 40 20
20
0 0
5
10 15 20 Variation coefficient (%)
Pb-212
Pb-214
Pu-238
25
0
30
5
10
15
20
25
Variation coefficient (%)
Ra-226
Pu-239
Ra-228
Ru-106
Sb-125
100
100
Sr-89: Sr-90: Th-232: Th-234:
60 40
6 89 93 17
20
Frequency (%)
80
80 Frequency (%)
Co-60: 239 40 20
20
0
Bi-214: 24 Ce-144: 10 Co-57: 15
60
U-234: 23 U-238: 27 U-nat: 295 Zn-65: 162
60 40 20
0
0 Variation coefficient (%)
Sr-89
Sr-90
Th
Th-234
0
5
10 15 Variation coefficient (%)
U-234
U-238
U-nat
20
25
Zn-65
Fig. 1. Uncertainty distributions of the measurements of Brazilian Laboratories analyzing low level radionuclides in different types of environmental samples.
411
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
analysis, the range of activity per unit mass or volume, the range of the reference value and its correspondent uncertainty were described. The results of analyses were classified using the normalized standard deviation, D; defined as the difference between the laboratory mean % and value of three independent determination results, X; the reference value, U; normalized to the standard deviation of the reference value, sU ; divided by the square root of the number of independent determinations n: Then D¼
ðX% UÞ pffiffiffi : sU = n
When the value of D is within 72 the laboratory have Good performance. The results fall in the intervals of 3oDo 2 or þ2oDo þ 3 are considered to be Acceptable, and for data with Dp 3 or DX þ 3 are considered as Out of control and the performance is not acceptable. The precision of each analysis can expressed by the variation coefficient for each type of radionuclide analyses, s VC ¼ % X%
where X% is the mean value of 3 independent determinations and s is the correspondent standard deviation. The quality criteria to determine good laboratory performance was the variation coefficient less than 5% in intercomparison runs. Results between 5% and 10% are also considered to be acceptable, for special types of matrices and radionuclides.
3. Results and discussion Samples were prepared with activity range per unit mass from 0.37 to 2.22 Bq/kg, except for 3H, from 111 to 555 Bq/kg and distributed by LNMRI. Samples with milk, vegetation and soil matrices were distributed in technical cooperation with Environmental Protection Agency, E.P. A and Environmental Measurements Laboratory, EML, of the United States. The activity range per unit mass of these samples are variable, but less than 1 kBq/kg, except for 137Cs and 40K. Fig. 1 shows the radionuclides analyzed in the intercomparison program, the activity range per unit mass for samples prepared and distributed by LNMRI, the number of analyses of each radionuclide and the occurrence frequency in each percentual range of
Table 1 Laboratories classified using accuracy and precision criteria, by radionuclide analyzed and number of participation intercomparison runs, with respective type of sample matrices Nuclide
Lab.
Intercomparison participation and sample matrix Water
Am-241 Ba-133
Co-60
LC AX BC CG DY EF EK NB PI QZ RA ZT AX CG DY EF EK LC PI QZ RA ZT
Milk
Air filter
Soil
Vegetation
1
4
4
11 6 17 9 11 8 24 22 7 16 20 11 22 9 13 8 24 7 16 22
1
1
3 1
1 1
6
4 1
1
6 6
4
6 6
D%
VC (%)
Good accuracy
Precision (%) o5
0.15 0.58 0.05 0.41 1.11 0.62 0.90 1.06 0.69 1.55 0.26 1.04
9.93 3.96 4.51 1.97 4.29 2.61 9.38 3.50 5.80 6.29 2.89 2.32
X X X X X X X X X X X X
1.07 0.05 0.81 0.37 0.94 0.35 0.44 1.98 0.28 0.20
5.45 2.13 4.35 2.78 9.31 4.62 4.75 5.42 3.59 2.14
X X X X X X X X X X
o10 X
X X X X X X X X X X X X X X X X X X X X X
412
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
Table 1 (continued) Nuclide
Lab.
Intercomparison participation and sample matrix Water
Cs-134
Cs-137
Gross alpha
Gross beta
H-3
Pb-210
Ra-226
AX BC CG DY EF NB PI RA ZT
11 6 21 9 12 25 24 16 21
AX BC CG DY EF FB LC PI QZ RA SP ZT
11 6 21 9 12 2 24 7 16 5 21
BC NB RA QZ ZT
6 4 14 11 18
DY QZ RA SK
9 11 14 24
NB PI RA ZT LC PI RA
27 24 16 17
AX EF EK LC RA SL
14 6 7 18 15 8
AXa CG EK LC LG MCa
29 7 8 26 26 38
Milk
Air filter
Soil
D%
VC (%)
Good accuracy
Precision (%) o5
Vegetation 1.13 0.37 0.05 0.80 0.35 0.23 0.47 0.27 0.84
7.12 5.16 2.37 4.44 2.59 5.42 5.65 4.79 3.40
X X X X X X X X X
0.47 1.24 0.38 0.04 0.13 0.22 0.68 0.49 1.48 0.73 0.51 0.54
4.87 4.06 2.41 3.87 4.99 1.84 1.85 4.52 6.42 3.03 4.28 1.54
X X X X X X X X X X X X
7 1 2
1.15 1.63 0.01 0.12 0.67
9.56 6.77 8.02 8.17 6.30
X X X X X
1 2 1 7
1.83 0.30 0.28 1.12
5.11 4.16 5.1 9.03
X X X X
0.32 0.43 1.32 0.79 0.05 0.78 1.03
3.37 4.75 4.18 4.54 4.81 1.97 5.92
X X X X X X X
X X X X X X X
1.38 0.31 0.06 1.00 0.36 1.98
7.79 3.01 2.88 1.94 4.76 3.82
X X X X X X
X X X X X
0.01 0.43 0.19 0.47 0.90 0.14
6.37 6.29 3.75 1.74 4.6 6.99
X X X X X X
1 1 2 1 1 3
1
1 1 2 3 1
7
4 6 4
2 1 3 6 1 2 7 7
1
7 4 7
6
1
6 2 6 6
6
6
7 7 7
1
1 1
6 6 6
o10 X X
X X X X X X X X X X X X X X X X X X X X X X X X X X X X
X
X X X X X X
413
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414 Table 1 (continued) Nuclide
Lab.
Intercomparison participation and sample matrix Water
Precision (%) o5
Vegetation X X
X
7 23 13 10
1.35 0.65 1.80 1.59
3.74 4.54 9.39 9.0
X X X X
X X
AX BC CG DY EF EK NB PI RA ZT
9 6 17 9 11 8 25 22 16 20
1.58 0.33 0.16 1.18 1.65 1.63 1.38 1.03 0.19 0.76
5.97 9.63 2.20 8.48 4.79 6.40 5.42 7.52 6.52 2.67
X X X X X X X X X X
X
Sr-90
RA
16
0.55
3.91
X
X
Th-232
AC EK LC MC OV RA
9 7 13 13 10 13
0.68 0.89 1.11 0.29 0.76 0.27
3.15 3.69 2.27 3.99 1.54 2.88
X X X X X X
X X X X X X
U-234
LC
4
6
0.17
5.95
X
U-238
LC SL
3
6 6
0.05 1.58
8.71 3.62
X X
0.90 0.33 0.21 0.54 1.17 0.22 0.42 0.89 0.73 0.13
2.87 6.58 3.71 2.36 5.33 1.39 2.95 8.58 2.6 3.41
X X X X X X X X X X
0.01 0.03 1.05 0.82 1.23 0.57 0.21 0.28 0.44
4.78 2.15 5.28 4.86 9.07 5.14 6.41 4.22 1.70
X X X X X X X X X
U-nat
Zn-65
a
CG LC LG QZ
Soil
Good accuracy
6.03 6.94
Ru-106
16 9
Air filter
VC (%)
0.18 1.02
Ra-228
RA SL
Milk
D%
AC AX EF LC MC OV RA SK SP ZT
12 31 22 27 10 16 22 10 22
BC CG DY EF EK NB PI RA ZT
6 17 9 11 8 25 22 16 20
1
1
1 1
3
3 1 5 3 4
6
6 3 7 5 7
6
o10 X
X X X X X X X X X X X
X X X X X X X X X X X X X X X X X X X X X
Laboratories which present intercomparison participation above 28 have requested additional samples for comparing different analytical techniques.
414
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
precision measurements. The best performance were for Th with 83.7% of the reported values having uncertainty less than 5%, 3H with 78.8%, and gamma emitters radionuclide determinations in water with 73% as mean value. The poor results were associated to 241 Am in soil and vegetation matrices, with 20% of results included in the quality indicator and 32% greater than 20%. Applying the quality criteria to the results of 3013 analyses obtained from 28 intercomparison runs performed from 1991 to 2000 for K, Th, U-nat, gross alpha, gross beta, and 27 other radionuclides in matrices of water (75.9%), soil (11.4%), vegetation (4.4%), air filter (7.7%) and milk (0.6%), we obtain the final results shows in Table 1. The laboratories participant in PNI are represented by its code name: AC, AX, BC, CG, CX, DY, EF, EK, FB, GF, LC, LG, MC, NB, OV, PI, QZ, RA, RL, SK, SL, SP, TW, ZT.
4. Conclusions These results suggest a detail examination and reduction of the uncertainty sources in measurement procedures and show the need of standardization of accreditation criteria of laboratory analyses. As the methodology used to evaluate the results of laboratories measurements depends on the reference value and its uncertainty for each radionuclide and its traceability to absolute standards and to BIPM, this type of intercomparison program contribute to establish the metrological network in the country.
References Jarvis, A.N., Siu, L., 1981. Environmental Radioactivity Laboratory Intercomparison Studies Program, EPA-600/481-004, EMSL-USEPA, Las Vegas.
Metrological capability of the Brazilian laboratories of analyses of radionuclides in environmental samples L. Tauhata*, M.E.C. Vianna, A.E. de Oliveira, A.C.M. Ferreira, C.C.S. da Conceic-a* o ! Laboratorio Nacional de Metrologia das Radiac-oes * Ionizantes (LNMRI), Instituto de Radioprotec-ao * e Dosimetria (IRD), Comissao * Nacional de Energia Nuclear (CNEN) Av. Salvador Allende, s/n Recreio dos Bandeirantes - Rio de Janeiro C.E.P. 22780-160, Brazil Accepted 30 July 2001
Abstract The metrological capability of 24 Brazilian laboratories of radionuclide determinations in environmental samples was analysed for purposes to establish requirements for accreditation of laboratory for each radionuclide. For each type of analysis, the range of activity per unit mass or volume, the range of the reference value and its correspondent uncertainty were described. The accuracy was established using the statistical criteria of USE.P.A. for the laboratory performance classification. The precision of the analyses, expressed by its variation coefficient of the laboratory determinations, was classified in intervals of 5%. The data base was constituted by 3013 results of analyses obtained in 28 intercomparison runs performed from 1991 to 2000, for K, Th, Unat, gross alpha, gross beta, and other 27 radionuclides, in matrices of water, soil, vegetation, air filter and milk. The quality criteria established were the variation coefficient less than 5% and good performance in intercomparison runs. Results included from 5% to 10% are also considered, for special types of matrices and radionuclides. r 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction The increasing of environmental samples analyses due to the exigency of quality control programs and studies of the environmental impact in many types of the human activities, claim to a fast response of the laboratories and the quality of the results assured by normative or accreditation organisms. The laboratories need to present results with accuracy and precision met recommendations of this type of institutions and, moreover, to submit its technical procedures for accreditation process. In the case of the laboratory analyses of radionuclide, it is important to note the necessity of an accreditation process for each kind of radionuclide. Large efforts are made in the past time to develop a reasonable quality *Corresponding author. Tel.: +55-2134-118185; fax: +552144-21605. E-mail address: [email protected] (L. Tauhata).
criteria to be applied in the classification of the laboratories and to evaluate its measurement performance. In the case of the laboratories that analyzing environmental samples the repeatability of measurements and their normalized standard deviation are seldom-included in performance evaluation, which should be included to study the long term stability of a laboratory performance. Then, one good criteria that can be obtained is to select the group that has the best performance over long period or many intercomparison runs. Another important requirement is related to the precision of the measurements, expressed by the variation coefficient. The low value of variation coefficient is, obviously, related to the low uncertainties of the results. By combining these two criteria, it is possible to establish the requirement to select the laboratories with confiability of measurement, for each radionuclide. Besides, it is important to take account of the main objective of the
0969-8043/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 9 - 8 0 4 3 ( 0 1 ) 0 0 2 2 3 - 8
410
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
measurement, in respect of the precision need. Other exigencies, such as the difficulties of doing the analysis, the type of sample matrix, type of radiation of emitter, and measurement procedure should be also considered. Another important point is to assure the traceability of the laboratories measurements to the national and international metrology network and to the BIPM.
2. Methodology To evaluate the performance in terms of accuracy of 24 Brazilian laboratories, the statistical criteria of Environmental Protection Agency of United States, USE.P. A, was used for the laboratory performance classification (Jarvis and Siu, 1981). For each type of 100
100
80
Ac-228 : 24 Am-241: 25 Ba-133: 162 Bi-212: 18
60 40
Frequency (%)
Frequency (%)
80
0
0
0
5
10 15 20 Variation Coefficient (%)
Ac-228
Am-241
25
Ba-133
0
30
Ce-144
Co-57
20
25
Co-60
100 80 Cs-134: 187 Cs-137: 316 Gross alpha: 195 Gross beta: 225
60 40
Frequency (%)
Frequency (%)
10 15 Variation Coefficient (%)
Bi-214
80
20
H-3: 99 K-40: 71 Mn-54: 15 Pb-210: 97
60 40 20
0
0 0
5
Cs-134
10 15 Variation Coefficient (%)
Cs-137
20
Gross alpha
25
0
5
Gross beta
10 15 Variation Coefficient (%)
H-3
K-40
Mn-54
20
25
Pb-210
100
80
Pb-212: Pb-214: Pu-238: Pu-239:
60 40
Frequency (%)
100
Frequency (%
5
Bi-212
100
21 23 6 27
80 Ra-226: 186 Ra-228: 139 Ru-106: 166 Sb-125: 11
60 40 20
20
0 0
5
10 15 20 Variation coefficient (%)
Pb-212
Pb-214
Pu-238
25
0
30
5
10
15
20
25
Variation coefficient (%)
Ra-226
Pu-239
Ra-228
Ru-106
Sb-125
100
100
Sr-89: Sr-90: Th-232: Th-234:
60 40
6 89 93 17
20
Frequency (%)
80
80 Frequency (%)
Co-60: 239 40 20
20
0
Bi-214: 24 Ce-144: 10 Co-57: 15
60
U-234: 23 U-238: 27 U-nat: 295 Zn-65: 162
60 40 20
0
0 Variation coefficient (%)
Sr-89
Sr-90
Th
Th-234
0
5
10 15 Variation coefficient (%)
U-234
U-238
U-nat
20
25
Zn-65
Fig. 1. Uncertainty distributions of the measurements of Brazilian Laboratories analyzing low level radionuclides in different types of environmental samples.
411
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
analysis, the range of activity per unit mass or volume, the range of the reference value and its correspondent uncertainty were described. The results of analyses were classified using the normalized standard deviation, D; defined as the difference between the laboratory mean % and value of three independent determination results, X; the reference value, U; normalized to the standard deviation of the reference value, sU ; divided by the square root of the number of independent determinations n: Then D¼
ðX% UÞ pffiffiffi : sU = n
When the value of D is within 72 the laboratory have Good performance. The results fall in the intervals of 3oDo 2 or þ2oDo þ 3 are considered to be Acceptable, and for data with Dp 3 or DX þ 3 are considered as Out of control and the performance is not acceptable. The precision of each analysis can expressed by the variation coefficient for each type of radionuclide analyses, s VC ¼ % X%
where X% is the mean value of 3 independent determinations and s is the correspondent standard deviation. The quality criteria to determine good laboratory performance was the variation coefficient less than 5% in intercomparison runs. Results between 5% and 10% are also considered to be acceptable, for special types of matrices and radionuclides.
3. Results and discussion Samples were prepared with activity range per unit mass from 0.37 to 2.22 Bq/kg, except for 3H, from 111 to 555 Bq/kg and distributed by LNMRI. Samples with milk, vegetation and soil matrices were distributed in technical cooperation with Environmental Protection Agency, E.P. A and Environmental Measurements Laboratory, EML, of the United States. The activity range per unit mass of these samples are variable, but less than 1 kBq/kg, except for 137Cs and 40K. Fig. 1 shows the radionuclides analyzed in the intercomparison program, the activity range per unit mass for samples prepared and distributed by LNMRI, the number of analyses of each radionuclide and the occurrence frequency in each percentual range of
Table 1 Laboratories classified using accuracy and precision criteria, by radionuclide analyzed and number of participation intercomparison runs, with respective type of sample matrices Nuclide
Lab.
Intercomparison participation and sample matrix Water
Am-241 Ba-133
Co-60
LC AX BC CG DY EF EK NB PI QZ RA ZT AX CG DY EF EK LC PI QZ RA ZT
Milk
Air filter
Soil
Vegetation
1
4
4
11 6 17 9 11 8 24 22 7 16 20 11 22 9 13 8 24 7 16 22
1
1
3 1
1 1
6
4 1
1
6 6
4
6 6
D%
VC (%)
Good accuracy
Precision (%) o5
0.15 0.58 0.05 0.41 1.11 0.62 0.90 1.06 0.69 1.55 0.26 1.04
9.93 3.96 4.51 1.97 4.29 2.61 9.38 3.50 5.80 6.29 2.89 2.32
X X X X X X X X X X X X
1.07 0.05 0.81 0.37 0.94 0.35 0.44 1.98 0.28 0.20
5.45 2.13 4.35 2.78 9.31 4.62 4.75 5.42 3.59 2.14
X X X X X X X X X X
o10 X
X X X X X X X X X X X X X X X X X X X X X
412
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
Table 1 (continued) Nuclide
Lab.
Intercomparison participation and sample matrix Water
Cs-134
Cs-137
Gross alpha
Gross beta
H-3
Pb-210
Ra-226
AX BC CG DY EF NB PI RA ZT
11 6 21 9 12 25 24 16 21
AX BC CG DY EF FB LC PI QZ RA SP ZT
11 6 21 9 12 2 24 7 16 5 21
BC NB RA QZ ZT
6 4 14 11 18
DY QZ RA SK
9 11 14 24
NB PI RA ZT LC PI RA
27 24 16 17
AX EF EK LC RA SL
14 6 7 18 15 8
AXa CG EK LC LG MCa
29 7 8 26 26 38
Milk
Air filter
Soil
D%
VC (%)
Good accuracy
Precision (%) o5
Vegetation 1.13 0.37 0.05 0.80 0.35 0.23 0.47 0.27 0.84
7.12 5.16 2.37 4.44 2.59 5.42 5.65 4.79 3.40
X X X X X X X X X
0.47 1.24 0.38 0.04 0.13 0.22 0.68 0.49 1.48 0.73 0.51 0.54
4.87 4.06 2.41 3.87 4.99 1.84 1.85 4.52 6.42 3.03 4.28 1.54
X X X X X X X X X X X X
7 1 2
1.15 1.63 0.01 0.12 0.67
9.56 6.77 8.02 8.17 6.30
X X X X X
1 2 1 7
1.83 0.30 0.28 1.12
5.11 4.16 5.1 9.03
X X X X
0.32 0.43 1.32 0.79 0.05 0.78 1.03
3.37 4.75 4.18 4.54 4.81 1.97 5.92
X X X X X X X
X X X X X X X
1.38 0.31 0.06 1.00 0.36 1.98
7.79 3.01 2.88 1.94 4.76 3.82
X X X X X X
X X X X X
0.01 0.43 0.19 0.47 0.90 0.14
6.37 6.29 3.75 1.74 4.6 6.99
X X X X X X
1 1 2 1 1 3
1
1 1 2 3 1
7
4 6 4
2 1 3 6 1 2 7 7
1
7 4 7
6
1
6 2 6 6
6
6
7 7 7
1
1 1
6 6 6
o10 X X
X X X X X X X X X X X X X X X X X X X X X X X X X X X X
X
X X X X X X
413
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414 Table 1 (continued) Nuclide
Lab.
Intercomparison participation and sample matrix Water
Precision (%) o5
Vegetation X X
X
7 23 13 10
1.35 0.65 1.80 1.59
3.74 4.54 9.39 9.0
X X X X
X X
AX BC CG DY EF EK NB PI RA ZT
9 6 17 9 11 8 25 22 16 20
1.58 0.33 0.16 1.18 1.65 1.63 1.38 1.03 0.19 0.76
5.97 9.63 2.20 8.48 4.79 6.40 5.42 7.52 6.52 2.67
X X X X X X X X X X
X
Sr-90
RA
16
0.55
3.91
X
X
Th-232
AC EK LC MC OV RA
9 7 13 13 10 13
0.68 0.89 1.11 0.29 0.76 0.27
3.15 3.69 2.27 3.99 1.54 2.88
X X X X X X
X X X X X X
U-234
LC
4
6
0.17
5.95
X
U-238
LC SL
3
6 6
0.05 1.58
8.71 3.62
X X
0.90 0.33 0.21 0.54 1.17 0.22 0.42 0.89 0.73 0.13
2.87 6.58 3.71 2.36 5.33 1.39 2.95 8.58 2.6 3.41
X X X X X X X X X X
0.01 0.03 1.05 0.82 1.23 0.57 0.21 0.28 0.44
4.78 2.15 5.28 4.86 9.07 5.14 6.41 4.22 1.70
X X X X X X X X X
U-nat
Zn-65
a
CG LC LG QZ
Soil
Good accuracy
6.03 6.94
Ru-106
16 9
Air filter
VC (%)
0.18 1.02
Ra-228
RA SL
Milk
D%
AC AX EF LC MC OV RA SK SP ZT
12 31 22 27 10 16 22 10 22
BC CG DY EF EK NB PI RA ZT
6 17 9 11 8 25 22 16 20
1
1
1 1
3
3 1 5 3 4
6
6 3 7 5 7
6
o10 X
X X X X X X X X X X X
X X X X X X X X X X X X X X X X X X X X X
Laboratories which present intercomparison participation above 28 have requested additional samples for comparing different analytical techniques.
414
L. Tauhata et al. / Applied Radiation and Isotopes 56 (2002) 409–414
precision measurements. The best performance were for Th with 83.7% of the reported values having uncertainty less than 5%, 3H with 78.8%, and gamma emitters radionuclide determinations in water with 73% as mean value. The poor results were associated to 241 Am in soil and vegetation matrices, with 20% of results included in the quality indicator and 32% greater than 20%. Applying the quality criteria to the results of 3013 analyses obtained from 28 intercomparison runs performed from 1991 to 2000 for K, Th, U-nat, gross alpha, gross beta, and 27 other radionuclides in matrices of water (75.9%), soil (11.4%), vegetation (4.4%), air filter (7.7%) and milk (0.6%), we obtain the final results shows in Table 1. The laboratories participant in PNI are represented by its code name: AC, AX, BC, CG, CX, DY, EF, EK, FB, GF, LC, LG, MC, NB, OV, PI, QZ, RA, RL, SK, SL, SP, TW, ZT.
4. Conclusions These results suggest a detail examination and reduction of the uncertainty sources in measurement procedures and show the need of standardization of accreditation criteria of laboratory analyses. As the methodology used to evaluate the results of laboratories measurements depends on the reference value and its uncertainty for each radionuclide and its traceability to absolute standards and to BIPM, this type of intercomparison program contribute to establish the metrological network in the country.
References Jarvis, A.N., Siu, L., 1981. Environmental Radioactivity Laboratory Intercomparison Studies Program, EPA-600/481-004, EMSL-USEPA, Las Vegas.