Intercomparison of TLD Services in Slovenia
Matjaž ŠTUHEC and Gorazd LAKOVIČ
Jožef Stefan Institute , Jamova 39, 1001 Ljubljana, Slovenia
Abstract
Five personal-dosimetry services from Slovenia and neighbouring countries were
examined, in order to intercompare the quality of their TLD systems. Irradiations were
carried out with different irradiation conditions, where dose, energy and angle of
incidence were varied. Comparison of the results with the ICRU performance test
requirements showed a rather good agreement within the prescribed trumpet curve, with
a few outliers on the upper edge or a few percent above the edge. An attempt was made
to analyse the results according to the IEC type test. Although the statistics doesn't
satisfy for decisive conclusions, the results indicate, that the investigated types of
dosimeters, with the right energy compensation, tend to comply with the tested
requirements.
Introduction
The quality of a supplier of personal-dosimetry service depends on the technical
characteristics of the approved dosimetry system and the training and experience of the
staff, together with the calibration procedures and quality assurance programmes. In
Slovenia procedures for approving the dosimetry systems at the regulatory authority
level are not yet established. As a preliminary investigation of performance quality of
the dosimetry systems, the intercomparison was organised by the Slovenian National
Laboratory for Dosimetry Standards. All three dosimetry services from Slovenia, one
service from Ruđer Bošković Institute, Croatia and one from Bosnia and Herzegovina
took part in the intercomparison exercise. In the first part a general performance test
was made according to the ISO 14146 [1] recommendations. In the second part, the
measured characteristics of four different dosimeter types were compared with the most
important IEC1066 [2] type-test requirements: linearity, energy dependence and angular
dependence.
Irradiations
The dosimeters were irradiated as sets of three on the ICRU slab water phantom
with Hp(10) doses. All irradiation conditions complied at the ISO 4037 [3] standard.
Overall uncertainty of the irradiation doses was 5%. Each group was irradiated with
specified conditions for dose, energy and angle of incidence, no fields with mixed
energies were used.
1.) Linearity was tested with three doses at Cs-137 in the range from 0.1 mSv to 5.5
mSv. The lower limit was chosen to be near the monthly recording level and the upper
limit in the range of annual limit for monthly monitoring, recommended by the ICRP
[4].
2.) Energy dependence was tested with 660 keV Cs-137, 1.2 MeV Co-60 and with Xrays at 40, 60 and 100 kV tube voltages with the narrow spectra qualities N40, N60 and
N100.
3.) For isotropy test, the dosimeters were irradiated with the N60 X-rays at 300 and 600
angels of rotation around vertical axis. Irradiations were split into two half time shots
for plus and minus angles, to get symmetrical conditions for the dosimeters lying out of
rotation axe.
For the last two tests, the doses were around 0.5 mSv. The dosimeters were then sent for
evaluation to participating laboratories. Unirradiated dosimeters were included to
measure travelling background dose, which was accounted for in the evaluation of the
results. Responses of the participating TLD systems under the defined conditions are
reported and compared to the directives of international standards.
Results and Discussion
After receiving the measured doses from the participants, and subtracting the
travelling dose, the response was calculated for individual dosimeter, as a ratio of
measured with true dose, R=Dm/Dt. In the first test, responses of all dosimeters were
compared with the so called ICRP trumpet curve [3].
Trumpet
2.0
R
1.5
1.0
0.5
0.0
100
1000
10000
dose
Fig.1: Responses R in comparison with the ICRP trumpet.
Figure 1 shows the responses for all the individual dosemeters as measured by the
participants in comparison with the ICRP trumpet. No additional corrections were made
to the evaluated doses at this stage of evaluation. Except for a few outliers, the
responses comply with the requirements of so called performance test for personaldosimetry services.
In the second stage of evaluation, an attempt was made to compare the
measurement capabilities of four participant systems, having different technical
characteristics, to the requirements of the IEC1066 type-test. Four different types of
personal dosimetry systems were compared: Harshaw and Rados systems with the
dosimeters based on LiF:Mg, Panasonic system with the combination of LiB and CaSO4
tablets in dosimeter badges, and IJS CaF2:Mg energy compensated dosimeters.
Although the statistical sample of the measurements, where only three dosimeters for
individual irradiation condition were used, doesn't allow conclusive judgements, we
tried to get at least an indication of how good the performance can be in comparison
with the standard. The results are in the table 1 for energy dependence and table 2 for
isotropy and linearity.
Table 1: Energy dependence - errors of measured doses as R-1 for systems I-IV.
E(keV)
I
II
III
IV
33
0.154
0.318
0.169
0.190
48
-0.108
0.066
-0.103
-0.010
83
-0.270
-0.218
0.038
-0.170
660
0.032
-0.065
0.015
0.050
1200
0.052
-0.040
0.015
0.033
Table 2: Isotropy - errors as R/R0-1, and linearity - errors as R-1.

I
II
III
IV
300
0.024
0.023
0.253
0.018
600
0.169
0.085
0.350
0.132
Sv
I
II
III
IV
105
0.007
0.013
0.005
0.043
530
0.032
-0.065
0.015
0.050
5554
0.032
-0.017
0.039
0.059
As the result of energy dependence and linearity, for each participant (I-IV) and each
irradiation condition, the relative error is calculated as R-1. R is the response of the
average of three measured doses for given irradiation, compared to the true dose,
<Dm>3/Dt. For isotropy, the relative response R/R0 is evaluated, that is the ratio of
response for angular incidence to the normal incidence. The results for energy
dependence are displayed in the figures 2,3 and the results for linearity and isotropy in
figures 4,5. Error bars in the figures indicate the measurement uncertainty for the groups
of three dosimeters, calculated as weighted standard deviation of the mean, for the
confidence interval of 95% probability.
0.300
0.225
0.150
R-1
0.075
0.000
33
-0.075
48
83
660
1200
-0.150
-0.225
-0.300
E(keV)
Fig.2: Energy dependence of the TLD systems I(red) and IV(blue).
0.270
R-1
0.180
0.090
0.000
-0.090
33
48
83
660
1200
-0.180
-0.270
E(keV)
Fig.3: Energy dependence of the TLD systems II(pink) and III(green).
0.210
0.140
R-1
0.070
0.000
30
60
`
105
530
5554
-0.070
-0.140
-0.210
X: 30/60deg
Cs: 0.1 - 5.5 mSv
Fig.4: Isotropy and linearity of the TLD systems I(red) and IV(blue).
0.350
0.250
R-1
0.150
0.050
-0.050
30
60
105
530
5554
-0.150
X: 30/60deg
Cs: 0.1 - 5.5 mSv
Fig.5: Isotropy and linearity of the TLD systems II(pink) and III(green).
Conclusions
The following conclusions can be drawn from the results for the individual types
of TLD systems:
I: The dosimeters have no energy compensation, measurement of energy dependence
results an average energy response of 1.58 for the X-ray energies. With this correction
factor taken into account as an average X-rays energy correction, we get the energy
dependence within the 30% IEC requirements. Isotropy complies at 300 incidence angle
and is just near above the edge of 15% requirements at 600. Linearity complies also at
the lowest dose, together with the broader, 10% confidence interval. A good batch
homogeneity is observed, with the overall spread of the individual readings in the
groups of three dosimeters to be 3% on the average, when leaving out a higher value of
8% only at the lowest dose.
II: The dosimeters are energy compensated, with the relative error for energy
dependence on the edge of 30% IEC requirements for the lowest energy X-rays, N40.
The measurements indicate, that isotropy complies with the 15% requirements, although
with a rather broad 10% confidence interval. Linearity seems to be within the 10%
requirements, except that the lowest dose shows a broader uncertainty because of an
outlier measurement. Average relative uncertainty of the groups of three dosimeters is
6% at high energies (leaving out the lowest dose) and about 10% for low energy
measurements.
III: The measurements tend to comply with the energy dependence requirements,
although with rather large 30% confidence interval for the N60, where the results is at
the limit for energy dependence. A good result is achieved for linearity, including the
lowest dose. On the other hand, a discrepancy is observed for isotropy at N60 quality Xrays, showing about 30% error, as well as measurement uncertainty. A further
investigation is needed with a better statistics to enlighten the findings.
IV: The measurements indicate an overall good result, showing tendency to comply
with all requirements, if better measurement statistics would be achieved. Poor
homogeneity of some of the tested groups seems to be the cause of the three outliers.
One of them, where one of the three dosimeter is obviously wrong at N60, is corrected
in the final result. In the rest of two, the doses of all three dosimeters are spread in about
20% confidence interval. Disregarding these outliers, a 4% overall average
measurement uncertainty is achieved.
Acknowledgments
Mr.B. Vekić is kindly acknowledged for his help at the stage of irradiations.
References
[1] International Organization for Standardization. Radiation Protection-Criteria and
Performance Limits for the Periodic Evaluation of Processors of Personal Dosimeters for
X and Gamma Radiation. ISO 14146, Geneva (2000).
[2] International Electrotechnical Commission. Thermoluminiscence dosimetry systems for
personal and environmental monitoring. IEC 1066 (1991).
[3] International Organization for Standardization. X and gamma reference radiation for
calibrating dosimeters and dose rate meters and for determining their response as a
function of photon energy. ISO/FDIS 4037 1-3 (1999).
[4] International Commission on Radiation Protection. General Principles for the Radiation
Protection of Workers Monitoring. ICRP 75 (1997).