Received 3 December 2016; revised 17 May 2017; editorial decision 21 June 2017; accepted 4 July 2017

Transcription

1 Radiation Protection Dosimetry (2018), Vol. 178, No. 2, pp Advance Access publication 15 July 2017 doi: /rpd/ncx091 NOTE COMPARISON OF PERSONAL DOSE EQUIVALENT Hp(10) IN 137 CS RADIATION BETWEEN THE PRIMARY STANDARDS LABORATORIES OF JAPAN AND AUSTRALIA USING BeO OSL PERSONAL DOSEMETERS D. J. Butler 1, *, T. Kurasawa 2, M. Litwin 1, J Mazaraki 1 and N. Saito 2 1 Australian Radiation Protection and Nuclear Safety Agency (ARPANSA), Yallambie, Victoria 3085, Australia 2 National Measurement Institute of Japan (NMIJ), AIST, Tsukuba Japan *Corresponding author: Duncan.Butler@arpansa.gov.au Received 3 December 2016; revised 17 May 2017; editorial decision 21 June 2017; accepted 4 July 2017 A comparison of personal dose equivalent Hp(10) for 137 Cs radiation was conducted between the primary standards laboratories of Japan and Australia. A set of 120 commercially available passive BeO OSL dosemeters were used (Dosimetrics GmbH, Munich). The aim was to investigate the precision which could be obtained with this technique, and to confirm the personal dose equivalent delivery methods in each standards laboratory. A dose of 5 msv was delivered to 40 dosemeters in each country, and 40 dosemeters were used as controls. The result of the comparison was a ratio of Hp(10) in Japan to Australia of with a combined standard uncertainty of 3.2%. The statistical uncertainty was 0.32% indicating that passive dosemeters can be used for comparisons of high precision. INTRODUCTION Primary and secondary standard dosimetry laboratories regularly perform comparisons to validate and establish equivalence of their standards. The majority of these comparisons are performed as a part of the CIPM Mutual Recognition Arrangement (CIPM MRA) (1). Under this arrangement Japan and Australia both reside in the Asia Pacific Metrology Programme (APMP) and have recently compared standards for air kerma and absorbed dose to water in regional and bilateral comparisons (2 4). The quantity of interest for personal monitoring services, however, is personal dose equivalent. The quantity Hp(d) is defined as the absorbed dose to water at d mm depth in an ICRU phantom in a specific irradiation geometry (5 7). For the purposes of this study we are concerned with Hp(10), the most common quantity used for personal monitoring for penetrating photon radiation. Only a few dosimetry laboratories have primary methods to determine Hp(10). Instead, it is common practice to establish the air kerma rate, free in air, and use a Monte Carlo calculation published in the international standard X and gamma reference radiation for calibrating dosemeters and doserate metres and for determining their response as a function of photon energy Parts 1 3:1996, 1997, 1999 (5 7), to calculate Hp(10). Personal dosemeters may then be irradiated on the surface of a phantom using the calculated Hp(10) value (7). Both ARPANSA and NMIJ provide irradiations for personal monitoring services based on air kerma standards and the application of ISO Hence, their values of Hp(10) should agree because their air kerma standards have been shown to agree to better than 1% (3). Nevertheless, a comparison was considered to be a worthwhile validation of the implementation of ISO 4037 at each laboratory. The Personal Radiation Monitoring Service (PRMS) at ARPANSA has recently changed from a CaS04:Dy TLD system to a commercially available BeO OSL system (Dosimetrics GmbH, Munich, 2015). The BeO OSL Personal Dosimeter meets the IEC Standard and has been type tested in Germany by Helmholtz-Zentrum dosimetry service (AWST Auswertungsstelle) and has been approved by the Physikalish Technischen Bundesanstalt (PTB). The new system has very small Type A uncertainties and the automated readout makes it easy to process large numbers of dosemeters. Therefore a high degree of precision should be possible if these dosemeters are used as the transfer instruments in a comparison. While other transfer instruments such as electronic personal dosemeters can be used (8 10), passive dosemeters are more representative of the majority of instruments used for personal monitoring. Passive dosemeters also make it much easier to conduct a blind comparison, which is important for proficiency testing and the accreditation of monitoring services. The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oup.com

2 PRMS is co-located with (although separate from) the standards laboratory at ARPANSA. Hence, an independent verification of its dosimetry is also of value, as it negates the potential for a claim of conflict of interest regarding the traceability chain. D. J. BUTLER ET AL. METHOD A group of 120 OSL personal dosemeters were prepared by the PRMS at ARPANSA, zeroed, and shipped to Japan. In Japan, 40 dosemeters were selected at random and irradiated to an Hp(10) of 5 msv, the dose being determined and delivered independently by NMIJ. The irradiation details are given in NMIJ irradiations. The dosemeters were returned to ARPANSA where they were received by the primary standards laboratory. Forty of the un-irradiated dosemeters were selected at random and given 5 msv, as determined by ARPANSA, leaving 40 to act as controls. All of the dosemeters were returned to PRMS and read out. The PRMS internal QC procedures confirmed that the readers were operating to specifications for the comparison readings. OSL personal dosemeters One of the personal dosemeters used for the comparison is shown in Figure 1. The personal dosemeters were shipped and irradiated inside blister packs (Figure 1a). The outer dimensions of the pack are mm 3. The light-tight black dosemeter (Figure 1b) has outer dimensions mm 3. The disassembled version (Figure 1c) shows the two active BeO areas. The area on the right is behind a thin window and can be used for Hp(0.07) assessments. For the comparison, only the readings from the active area on the left were used. The fading, linearity, energy dependence and Type A uncertainty associated with reading the dosemeters has already been established (11, 12). When using the dosemeters for a comparison, we decided to minimise the use of corrections by choosing dosemeters zeroed at the same time, and by delivering (as close as possible) the same dose at each laboratory. In reference (11) the fading of these dosemeters is established as being negligible after the first 10 min following irradiation, and so no correction was applied for the 2-week difference between the NMIJ and ARPANSA irradiations. NMIJ irradiations The NMIJ irradiation facility consists of a 34 TBq 137 Cs source with a collimator and iron plate attenuator to reduce the dose rate (Figure 2). A water-filled phantom was up at 400 cm distance from the source. The NMIJ phantom has outer dimensions 32 cm 32 cm cm and is made from PMMA walls 1 cm thick except the front face which is 0.25 cm thick. Figure 1. BeOSL personal dosemeter (Dosimetrics GmbH, Munich) (a) in blister pack, as worn by users and as used for the comparison, (b) removed from blister pack and (c) when removed from the light-tight package, showing the two active BeO elements. The air-kerma rate was determined by the NMIJ primary graphite cavity chamber (13). The irradiation time was about 630 s to deliver 5 msv. The centre of the personal dosemeter was set at the reference point (400 cm). The 2 mm PMMA plate was placed in front of personal dosemeters for the build-up plate. ARPANSA irradiations The ARPANSA irradiation facility consists of a 50 TBq 137 Cs source in a custom-made housing and collimator. A water-filled phantom was set up at 400 cm distance from the source (Figure 3). The ARPANSA phantom has outer dimensions 32 cm 32 cm 16.3 cm and is made from PMMA walls 1 cm thick except the front face which is 0.3 cm thick. The phantom is filled with water. Source timing is done manually. For short exposures, timing errors and source transit effects need to be taken into account. For this reason a transmission monitor chamber (Stealth ChamberTM, IBA Dosimetry GmbH, Schwarzenbruck, Germany) was 236

3 COMPARISON OF PERSONAL DOSE EQUIVALENT HP(10) Figure 2. Irradiation arrangement at the NMIJ. Figure 3. Irradiation arrangement at ARPANSA. positioned over the exit port of the collimator. The charge from the chamber was integrated during each exposure and used to determine the delivered dose. The monitor chamber was calibrated using a Farmer chamber positioned at the reference point (400 cm), free in air, in the absence of the phantom and buildup plate, using exposures of the same duration as required to deliver 5 msv (125 s). Traceability NMIJ maintains the Japanese primary standard for air kerma at 137 Cs and 60 Co (a graphite cavity chamber (13, 14) ). ARPANSA maintains the Australian primary standard for air kerma at 60 Co (a graphite cavity chamber (15) ) and for medium-energy kilovoltage X-rays (a free-air chamber (16) ). The response of a Farmer chamber to 137 Cs gamma rays was determined by linear interpolation between these standards (Table 1). This chamber was then used to determine the air kerma rate in the 137 Cs beam. RESULTS The results of the PRMS readouts of the 120 dosemeters are summarised in Table 2. The average reading of the un-irradiated dosemeters was msv. The average of the dosemeters irradiated in Japan was msv and those in Australia msv. Note that PRMS is traceable to ARPANSA and so the readouts for ARPANSA are expected to average exactly 5 msv above the control. For an individual dosemeter i, wedefine the ratio R i as the ratio of the dose delivered to the dosemeter according to the laboratory (D i,lab ), to the PRMS readout P i less the background signal (Eq. 1). Here the background, B, is determined by averaging the PRMS readouts of the control dosemeters (0.088 msv). R = D /( P B) ( 1) i i,lab i The average, standard deviation and standard deviation of the mean are given in Table 3. For the ARPANSA the average ratio was and for NMIJ 237

4 D. J. BUTLER ET AL. Table 1. Calibration coefficient and traceability for the Farmer chamber used to determine the air kerma rate in the ARPANSA 137 Cs beam. Description Traceability ARPANSA beam code Effective energy (kev) N K (mgy/nc) 60 Co gamma rays Primary standard cavity chamber S-Co kvp X-ray beam Free air chamber NXA Cs gamma rays Interpolation of above S-Cs Table 2. PRMS readouts from the OSL personal dosemeters. Irradiating laboratory Number of dosemeters Mean reading (msv) Standard deviation (msv) Standard deviation of the mean (msv) None (controls) NMIJ ARPANSA Table 3. Average ratios of the delivered dose to the reading, for ARPANSA and NMIJ, and the ratio of NMIJ to ARPANSA calculated from these ratios. Ratio between Number of dosemeters Mean of R i (delivered/reading) Standard deviation in R i (%) Standard deviation of the mean (%) ARPANSA/ PRMS NMIJ/PRMS Ratio of mean ratios NMIJ/ARPANSA a a The combined standard uncertainty evaluated from the Type A uncertainty only (see Table 6). it was 1.015, so that the overall comparison result was a ratio of NMIJ to ARPANSA of The combined statistical uncertainty in this ratio is 0.32%. The comparison result r can be expressed for a general case of n dosemeters irradiated at LAB1 and m dosemeters irradiated at LAB2, as follows: r = m n n Di P,LAB1 m Di,LAB2 B P B j j,lab2 i= 1 i,lab1 = 1 ( 2) where D i, LAB are the doses delivered to each dosemeter according to LAB1 or LAB2. The comparison ratio r is therefore a measure of the agreement between the dose determined by each laboratory. In the case where each laboratory delivers the same target dose, and the same number of dosemeters are used: r = n n 1 1 P i B P B j j i= 1,LAB1 = 1,LAB2 ( 3) Uncertainties Uncertainty budgets for the Hp(10) irradiations in each laboratory are given in Tables 4 and 5. NMIJmaintain a primary standard at 137 Cs, whereas ARPANSA interpolate the response of a Farmer 2571 ionisation chamber calibrated against their primary air kerma standards at 60 Co and medium-energy X-rays. The dominating uncertainty in each case is the conversion factor from air kerma to personal dose equivalent. This value (2% at k = 1) for ARPANSA is given in the ISO 4037 standard (7) and takes into account the different spectra and different geometries which are possible between different sources and using a range of source phantom distances. NMIJ measured a photon energy spectrum to determine this conversion factor using a Ge spectrometer. Then, the uncertainty of this value is 0.9% at k = 1, which is smaller than 2%. In practice this uncertainty is likely to be partially correlated between ARPANSA and NMIJ, especially since each laboratory used the same source detector distance. An additional uncertainty in the comparison ratio arises because of the different reference points used at NMIJ and ARPANSA. NMIJ used the centre 238

5 COMPARISON OF PERSONAL DOSE EQUIVALENT HP(10) Table 4. Uncertainty in the NMIJ irradiations. Quantity Value Standard uncertainty, u (%) Air kerma rate at 4000 mm Gy/s 0.4 Conversion from air kerma to Hp(10) on slab phantom 1.22 Sv/Gy 0.9 Beam uniformity over 10 cm irradiation area Uncertainty for the build-up plate Combined uncertainty in delivered Hp(10) 1.8 Table 5. Uncertainty in the ARPANSA irradiations. Quantity Value Standard uncertainty, u (%) N K of Farmer chamber for 137 Cs a Gy/C 1.2 Calibration of monitor chamber positioned Gy/C 0.5 b at exit port for air kerma at 4000 mm. Distance error between air kerma calibration 5 mm in 4000 mm 0.3 point and dosemeter reference point Beam uniformity over 10 cm irradiation area Timing/source transit error when used 0.1 s in 125 s irradiation 0.1 with monitor chamber Conversion from air kerma to Hp(10) on slab phantom 1.21 Sv/Gy 2.0 Combined uncertainty in delivered Hp(10) 2.6 a Details of the air kerma primary standard uncertainties for ARPANSA are given in references (15, 16). b The overall monitor chamber uncertainty consists of the Farmer chamber uncertainty added in quadrature with the additional uncertainty shown in this cell. Quantity Table 6. Uncertainty in the comparison ratio. Standard uncertainty in r (%) Statistical uncertainty in ratio 0.32 Reader stability (included in the above) Uncertainty due to background <0.1 subtraction a Fading between NMIJ and 0.20 ARPANSA irradiations (14 days) Combined uncertainty in r 0.40 a Includes reader linearity and statistical variation in the background and its readout. of the dosemeter, and ARPANSA used the rear surface. At 4 m the correction for the difference of 5 mm is 0.3%. In future ARPANSA will use the centre of the dosemeter, in accordance with ISO In Table 6 we present the uncertainties in the comparison ratio. The major contributor is the statistical variation in the dosemeter readouts, which yield a combined uncertainty of 0.32% in the ratio r (Table 3). This uncertainty includes the variation in the doses received by the dosemeters (from the laboratory irradiation and otherwise), and the variability of the reader over the course of the readout (about half an hour). It can be shown from equation three that the comparison ratio is insensitive to the background value, provided it issmall,andsocontributionstotheratiouncertainty from the statistical variation in the dosemeters used to evaluate the background, and any reader non-linearity associated with the subtraction of a small background from a larger dose reading, are negligible. In reference (11) the fading is shown to be ~92 96% (with no trend) over periods of days after irradiation, with a standard uncertainty on each measurement of around 2% (Figure 4 of reference (11) ). If we assume that fading, if present, is linear after the first day, and that it is not more than 4% over 300 days, then fading difference over the 14-day interval between the irradiations in our comparison should be <0.2%. DISCUSSION The control dosemeters received an average of msv with a standard deviation of msv. 239

6 The period between zeroing the dosemeters and reading them out was 30.6 days, so that this dose corresponds to 2.9 μsv/d (1.0 msv/y). This value is consistent with the external component of the natural background in Australia and Japan, and suggests that any additional exposure due to air travel or security screening was not significant. Dosemeters kept at ARPANSA can be expected to accumulate dose at the rate of ~2.3 μsv/d. The background is 1.8% of the 5 msv comparison dose. For high accuracy comparisons, then, the background must always be assessed. If smaller doses are used, or comparisons last longer than 30 days, this requirement becomes even more important. The ratio of ARPANSA to the backgroundcorrected PRMS reading was with a statistical uncertainty of 0.21%. Ideally the ratio would be unity because PRMS is traceable to ARPANSA. However, the difference of 0.9% is consistent with the combined statistical uncertainties of the comparison result and the PRMS system calibration. The value is also well within the acceptable limit for reader stability. The overall comparison result NMIJ/ARPANSA of has a combined standard uncertainty of 3.2%. The major contributions to this value are the statistical uncertainty of 0.32% and the standard uncertainty of each laboratory, 2.6% for ARPANSA and 1.8% for NMIJ. The result is therefore well within the expected uncertainty. That the agreement is significantly better than the combined uncertainty is probably an indication of a high level of correlation in the uncertainty for the conversion factor from air kerma to Hp(10), and should not be taken to imply that personal dose is determined at higher accuracy than the given uncertainty budgets. The purpose of the study was 3-fold: to determine the precision of a comparison using passive dosemeters, to provide validation for the implementation of ISO 4037 in each laboratory and to provide the PRMS with an independent verification of their dosimetry. We have shown that a high level of precision is possible, provided many dosemeters are used. The other aims of the study are satisfied by the close agreement found. The use of passive dosemeters is novel. Usually, international comparisons use electronic personal dosemeters or ion chambers (8 10). However, when many transfer instruments are used we have shown that a similar precision can be achieved. The advantage of using passive dosemeters is that the irradiation service is tested using the devices it is ultimately intended for. Hence, the results support the standards laboratories irradiations for the proficiency testing and accreditation of personal monitoring services. The potential disadvantages are the lack of immediate results, and the need to use many dosemeters in order to get acceptable precision. D. J. BUTLER ET AL. CONCLUSION A comparison of personal dose equivalent using passive personal dosemeters was successfully conducted. The ratio of Hp(10) delivered at ARPANSA and NMIJ was determined with a statistical uncertainty of 0.32%. The high precision is similar that achievable with more commonly used transfer instruments such as ionisation chambers. The ratio of Hp(10) between NMIJ (Japan) and ARPANSA (Australia) was found to be with a combined standard uncertainty of 3.2%. The control dosemeters received a dose consistent with the external source component of natural background, indicating that screening and travel did not add significantly to the dosemeter doses during the comparison. REFERENCES 1. The CIPM MRA (Mutual Recognition Arrangement of national measurement standards and of calibration and measurement certificates issued by national metrology institutes), first signed in Accessed online in 2016: pdf 2. Allisy-Roberts, P. J., Kessler, C. and Burns, D.T. Summary of the BIPM.RI(I)-K5 comparison for air kerma in 137 Cs gamma radiation. Metrologia 50(Tech. Suppl.), (2013). 3. Chun, K. J. et al. Final report on APMP.RI(I)-K1: APMP/TCRI key comparison report of measurement of air kerma for 60 Co gamma-rays. Metrologia 50(Tech. Suppl.), (2013). 4. Allisy-Roberts, P. J. and Burns, D. T. Summary of the BIPM.RI(I)-K4 comparison for absorbed dose to water in 60 Co gamma radiation. Metrologia 42(Tech. Suppl.), (2005). 5. ISO X and gamma reference radiation for calibrating dosemeters and doserate meters and for determining their response as a function of photon energy Part 1: Radiation characteristics and production methods (1996). 6. ISO X and gamma reference radiation for calibrating dosemeters and doserate meters and for determining their response as a function of photon energy Part 2: Dosimetry for radiation protection over the energy ranges from 8 kev to 1.3 MeV and 4 MeV to 9 MeV (1997). 7. ISO X and gamma reference radiation for calibrating dosemeters and doserate meters and for determining their response as a function of photon energy Part 3: Calibration of area and personal dosemeters and the measurement of their response as a function of energy and angle of incidence (1999). 8. Ankerhold, U., Tiefenböck, W., Witzani, J. and Ambrosi, P. Supplementary comparison: EUROMET.RI(I)-S1: personal dose equivalent comparison between the BEV and the PTB. Metrologia39(Tech. Suppl), (2002). 9. Ankerhold, U. and Hupe, O. Supplementary comparison: EURAMET supplementary comparison of the personal dose equivalent quantity for photon radiation: 240

7 COMPARISON OF PERSONAL DOSE EQUIVALENT HP(10) EURAMET.RI(I)-S5. Metrologia 49(Tech. Suppl), 0601 (2012). 10. IAEA-TECDOC Intercomparison of Personal Dose Equivalent Measurements by Active Personal Dosimeters (Vienna: IAEA) (2007). 11. Sommer, A. and Henniger, J. Investigation of a BeObased optically stimulated luminescence deosemeter. Radiat. Prot. Dosim. 119(1 4), (2006). 12. Sommer, M., Jahn, A. and Henniger, J. A new personal dosimetry system for Hp(10) and Hp(0.07) photon dose based on OSL-dosimetry of beryllium oxide. Radiati. Meas. 46, (2011). 13. Kessler, C., Saito, N. and Kurosawa, T. Key comparison BIPM.RI(I)-K5 of the air kerma standards of the NMIJ, Japan and the BIPM in 137 Cs gamma radiation. Metrologia 50(Tech. Suppl.), (2013). 14. Kessler, C., Saito, N. and Kurosawa, T. Key comparison BIPM.RI(I)- K1 of the air kerma standards of the NMIJ, Japan and the BIPM in Co-60 gamma radiation. Metrologia 50(Tech. Suppl.), (2013). 15. Kessler, C., Allisy-Roberts, P.J., Lye, J.E. and Oliver, C. Comparison of the standards for air kerma of the ARPANSA and the BIPM for 60 Co gamma radiation. Metrologia 48(Tech. Suppl.), (2011). 16. Burns, D.T., Lye, J.E., Roger, P. and Butler, D.J. Key comparison BIPM.RI(I)-K3 of the air-kerma standards of the ARPANSA, Australia and the BIPM in medium-energy x-rays.metrologia49(tech. Suppl.), (2012). 241