In Situ Gamma Spectrometry Intercomparison in Fukushima, Japan

Transcription

1 Jpn. J. Health Phys., 50 (3) (2015) Technical Data In Situ Gamma Spectrometry Intercomparison in Fukushima, Japan Satoshi MIKAMI,* 1 Shoji SATO,* 2 Yoshifumi HOSHIDE,* 3 Ryuichi SAKAMOTO,* 4 Naotoshi OKUDA* 3 and Kimiaki SAITO* 5 Received on May 1, 2015 Accepted on September 17, 2015 Intercomparison of in situ gamma spectrometry was organized at a site contaminated by the radioactive fallout that originated from the Fukushima Dai-ichi Nuclear Power Plant accident. This intercomparison was conducted by eight teams from four different institutions, which have contributed to the government-led project to construct distribution maps of radionuclides deposited on the ground soil. The resultant 134 Cs and 137 Cs inventories evaluated by the participants agreed within 6% of the 40 K inventories distribution maps of the radionuclide inventory in the ground soil. KEY WORDS: intercomparison, in situ, gamma spectrometry, Fukushima, radiocesium, natural radionuclides, inventory. I INTRODUCTION The radioactive fallout that originated from the Fukushima Dai-ichi Nuclear Power Plant accident (referred to as the accident ) following the Tohoku-Paci c Ocean Earthquake deposited onto the ground over a large area in east Japan. To create radionuclide distribution maps, in situ gamma spectrometry using a portable gamma spectrometer has been adopted to evaluate the inventory (activity per unit area for anthropogenic radionuclides and activity per unit mass for natural radionuclides) of radionuclides exist in the ground soil around the Fukushima Dai-ichi Nuclear Power Plant. 1 3) In situ gamma spectrometry can quantify average inventories over a large area around the point of measurement, resulting in less statistical uctuation of the obtained data than that obtained by soil-sampling-based evaluations. 4 9) Intercomparison is useful for verifying the reliability, accuracy, and consistency of measurements among the participants. In case of intercomparison of sample measurements among different laboratories, a common * 1 Fukushima Environmental Safety Center, Sector of Fukushima Research and Development, Japan Atomic Energy Agency; 7 1, Taira Omachi, Iwaki-shi, Fukushima , Japan. mikami.satoshi@jaea.go.jp * 2 Radioactive Analysis Division, Japan Chemical Analysis Center; 295 3, Sanno-cho, Inage, Chiba-shi, Chiba , Japan. * 3 Division of Emergency Preparedness Technology, Nuclear Safety Technology Center; , Hakusan, unkyo, Tokyo , Japan. * 4 Institute of Radiation Measurements; 2 4 Shirane, Shirakata, Tokaimura, Naka-gun, Ibaraki , Japan. * 5 Fukushima Environmental Safety Center, Sector of Fukushima Research and Development, Japan Atomic Energy Agency; 2 2 2, Uchisaiwaicho, Chiyoda, Tokyo , Japan. reference material can be measured at each laboratory. Conversely, eld measurements need to be performed at the same eld site to allow intercomparison. Subsequently, simultaneous or sequential measurements at the same position should be conducted within a short time span to avoid possible changes in environmental conditions or contaminant levels at the site. The authors have participated in eld measurements using portable gamma spectrometers in a 2013 measurement campaign. 10) Eight teams from four institutions participated in the campaign. To begin the campaign, an intercomparison for in situ gamma spectrometry was organized and simultaneous measurements were conducted at speci c locations with their respective instruments (detector systems) for con rming the measurement accuracy and demonstrating the compatibility among the teams. This paper presents the results of the intercomparison conducted at a site affected by the radioactive fallout due to the accident. II MATERIALS AND METHODS Two intercomparisons were conducted on June 3 and 10, 2013 at a site in the Fukushima Prefecture. Six teams from four institutions performed simultaneous measurements on the rst day. Three teams from one institution, including one team from the rst day, conducted measurements on the second day at the same site. The arrangement for the intercomparison on the rst day is shown in Fig Site description The intercomparisons were conducted at a non-paved playground surrounded by rice elds. This playground had not been decontaminated, and it was nearly at and

2 In Situ Gamma Spectrometry Intercomparison in Fukushima, Japan 183 Fig. 1 The arrangement of the measurement site for the intercomparison on the rst day. The photo was taken on June 3, 2013 facing south. suf ciently large for in situ spectrometry with an area of approximately m2. Portable gamma spectrometers were placed in concentric circles around a reference point. A three-story concrete building and a few low structures were approximately 90 m north and 60 m south from the reference point, respectively. The reference point for the intercomparison measurements was predetermined on the basis the position information of a former mapping project. 2, 3) On the rst day, the exact coordinates of the reference point were determined by a Global Positioning System (GPS). Three concentric circles with 3, 6, and 9 m radii centered around the reference point were drawn on the ground, as shown in Fig. 2. Four points were then marked at the vertices of the square inscribed in each circle and at the mid-point between the points on the outermost circle. On the second day, the reference point was set according to the coordinates obtained on the rst day. However, the reference points each day were not identical as the determined reference point from the rst day could not be marked on the ground. The reference point for the second intercomparison was reset as close to the same point as possible according to the coordinates recorded on the rst day, although the positioning accuracy of the GPS was approximately 20 m. Figure 3 indicates the positional relation of the points of measurement on the second day. There were 16 and 12 measurement points on the ground for the rst and second days, respectively. At each of these points, as well as the center point of each concentric circle, in situ gamma spectrometry and/or the measurements of ambient 0.72± ± ±0.01 F 0.59±0.01 B 0.60± ±0.01 A 0.66±0.01 E C 0.55± ± ± ±0.03 D 0.53± ± ± ± ± ±0.02 3m 6m 9m Fig. 2 Positions of in situ gamma spectrometry and dose rate measurements at the measurement site on the rst day. Circles represent the points of dose rate measurement. Arabic numerals show the dose rate in air 1 m above the ground obtained as the average with its standard deviation of ve readings in ȝsv h 1. Alphabetic letters represent the points of measurement and participant codes, for in situ gamma spectrometry.

3 184 Satoshi MIKAMI, Shoji SATO, Yoshifumi HOSHIDE, Ryuichi SAKAMOTO, Naotoshi OKUDA and Kimiaki SAITO 0.59± ± ± ± ± ±0.01 H B 0.64± ±0.01 G 0.58±0.01 Ta ble 1 Results of the dose rate uniformity measurement at the intercomparison site. area, in radius [m] Statistical parameter 1 ] Std. Deviation Coeff. Vari. [%] Number n Min Max ± ± ± ±0.01 Fig. 3 Positions of in situ gamma spectrometry and dose rate measurements at the measurement site on the second day. Circles represent the points of dose rate measurement. Arabic numerals show the dose rate in air 1 m above the ground obtained as the average with its standard deviation of ve readings in Sv h 1. Alphabetic letters represent the points of measurement and participant codes, for in situ gamma spectrometry. dose equivalent rate (hereinafter referred to as dose rate in air ) were conducted (Figs. 2 and 3). (1) Homogeneity of radionuclide deposition at the site To verify the homogeneity of radionuclide deposition (at this time, mainly radiocesium isotopes), the measurements of dose rate in air were conducted 1 m above the ground with an energy-compensated NaI(Tl) scintillation survey meter (TCS- 171 manufactured by Hitachi Aloka Medical Ltd.). The dose rate in air at each point was obtained as the average of ve readings in Sv h 1. This was done on the rst and second days at points 17 and 13, respectively, within a 9 m radius of each reference point. For the rst day, the dose rate in air ranged from 0.53 to 0.61 Sv h 1 and 0.53 to 0.95 Sv h 1 within a 3 and 9 m radius from the center, respectively (Fig. 2 and Table 1). For the second day, the dose rate in air ranged from 0.58 to 0.70 Sv h 1 and 0.58 to 0.75 Sv h 1 within a 3 and 9 m radius from the center, respectively (Fig. 3 and Table 1). For the rst day, a relatively higher dose rate than the neighboring areas was found further than 9 m southwest from the center. These relatively higher dose rate areas were found to correspond to the area where grass was thicker, i. e., the distribution of dose rate in air at the site was partially inhomogeneous. 3m 6m 9m 2. Instrumentation, measurement condition and con guration of intercomparison (1) Instrumentation Each participating team used a portable gamma spectrometer with a high-purity germanium detector mounted on a tripod, allowing the effective center of the detector to be 1 m above the surface of the ground. Two n-type and six p-type germanium detectors were employed by the participants. The speci cations of the instruments used for intercomparison are listed in Table 2. (2) Intercomparison 1 Six teams (A to F) were participated a simultaneous measurement on the rst day (June 3, 2013). The detectors were placed at the center of concentric circles, four points on the innermost circle (3 m radius) and one point on the 6 m radius circle northwest of the center (Fig. 2). Inventories of 134 Cs, 137 Cs, 110m Ag, 214 Pb, 214 i, 208 Tl, 228 Ac, and 40 K were evaluated by integrating the counts in the peak region corresponding to the energy of 604 kev, 662 kev, 885 kev, 352 kev, 1,765 kev, 583 kev, 911 kev, and 1,461 kev, respectively, in the gamma-ray spectrum, in accordance with ICRU Report 53. 5) The data acquisition time was 0.5 h in live time. For this intercomparison, a relaxation mass depth of 2.06 g cm 2 was adopted for the evaluation of radiocesium inventory, according to the on-site investigations of radiocesium distribution depth by the scraper plate method. 11) (3) Intercomparison 2 Three teams (, G, and H) were conducted simultaneous measurements on the second day (June 10, 2013). Any athletic activities and decontamination work were not conducted at the site between the rst and second intercomparisons, according to interviews with the site manager. The precipitation in the region between the intercomparison 1 and 2 was reported to be 18.5 mm, though it was no rain on both days. Evaluation method of radionuclides inventory, the measurement time and adopted relaxation mass depth were the same as those for the intercomparison 1. III Results and Discussion Cesium-134, 137 Cs, 208 Tl, 228 Ac, and 40 K were detected by all teams and 110m Ag was detected by one team with activity slightly over the limit of detection. Lead-214 and 214 i were detected by a few teams. The evaluated radionuclide inventories detected by all teams were used in the

4 In Situ Gamma Spectrometry Intercomparison in Fukushima, Japan 185 Table 2 Detector speci cations. intercomparison. To estimate measurement consistency, the evaluated inventories of 134 Cs, 137 Cs, 208 Tl, 228 Ac, and 40 K were compared for every nuclide. 1. Results for radiocesium The results of the intercomparison with respect to 134 Cs and 137 Cs are shown in Tables 3 6. Tables 3 and 4 statistically summarize the results over both days for 134 Cs and 137 Cs, respectively. As an example, the coef cients of variation (C. V., which is de ned as the ratio of the standard deviation to the mean.) of deposited 134 Cs inventories were approximately 7.4 and 9.6% without correction for inhomogeneous distribution on the rst and second days, respectively, as shown in Table 3. It was reported that radiocesium was the dominant contributor of the dose rate in air in the environment around the Fukushima site at the time of the study. 3) Therefore, it was considered that the inhomogeneous distribution of the dose rate in air, as reported above, was caused by the inhomogeneous deposition of radiocesium. Hence, comparison in terms of the radiocesium inventory corrected by the dose rate in air was also implemented. After correction for inhomogeneous distribution by normalization by the dose rate in air, the C. V. (1 sigma) with respect to the ratio of 134 Cs inventory to the dose rate in air was approximately 6% among the six teams on the rst day and approximately 5% among the three teams on the second. Those for 137 Cs were approximately 6% among the six teams on the rst day and approximately 4% among the three teams on the second, as shown in Tables 3 and 4. Table 5 shows percent differences from the group mean in terms of the ratio of 134 Cs or 137 Cs inventories to the dose rate in air measured by the survey meters where a portable gamma spectrometer was Table 3 Results of the intercomparison for 134 Cs. Statistical 134 Cs Dose Rate* 134 Cs/DoseRate* 134 Cs Dose Rate* 134 Cs/DoseRate* Parameter 2 ] 1 ] 2 1 ] 2 ] 1 ] 2 1 ] Mean Std. Devi Coeff. Vari.% Number n Min Max * Dose Rate: Ambient dose equivalent rate measured by a NaI(Tl) scintillation survey meter.

5 186 Satoshi MIKAMI, Shoji SATO, Yoshifumi HOSHIDE, Ryuichi SAKAMOTO, Naotoshi OKUDA and Kimiaki SAITO Table 4 Results of the intercomparison for 137 Cs. Statistical 137 Cs Dose Rate* 137 Cs/Dose Rate* 137 Cs Dose Rate* 137 Cs/Dose Rate* Parameter 2 ] 1 ] 2 1 ] 2 ] 1 ] 2 1 ] Mean Std. Devi Coeff. Vari.% Number n Min Max * Dose Rate: Ambient dose equivalent rate measured by a NaI(Tl) scintillation survey meter. Ta ble 5 Percent differences from the group mean for 134 Cs and 137 Cs from the intercomparison on June 3, Participant Code A C D E F ( 134 Cs/ dose rate) ( 137 Cs/ dose rate) placed on the rst day. Table 6 shows those on the second day. The data shows that the maximum percent differences from the group mean were 7.7% for 134 Cs and 8.5% for 137 Cs on the rst day. Those on the second day were 5.2% for 134 Cs and 3.8% for 137 Cs. Furthermore, the depicted intercomparison results for 134 Cs and 137 Cs are shown in Figs. 4 and 5, respectively. Relative uncertainties in measurement of 134 Cs and 137 Cs inventories normalized by the dose rate in air were estimated to be approximately 6% in maximum for both, for the rst day (coverage factor, k =2). 12) Those of 134 Cs and 137 Cs for the second day were 5% (k = 2) in maximum for both. We recognized that the results falling within 10% of C. V. are in good agreement in the eld having inhomogeneous distribution of dose rate in air, and those within 6% of C. V. after correction for the inhomogeneous distribution, to draw up a single map of radionuclide deposition by working together with eight teams. Uncertainty in inventory according to inhomogeneity of deposited radiocesium is considered to be large. These results are also useful to compare the results of a similar previous study that showed 14 19% of C. V. among 24 participating teams at a larger measurement area than that in this work, for example. 8) 2. Results for natural radionuclides The results of intercomparison with respect to natural radionuclides are shown in Tables 7 9. The C. V. for 40 K was approximately 4% among the six teams on the rst day and 2% among the three teams on the second day. The C. V. for 208 Tl was approximately 14% for each intercomparison, as shown in Table 7. Large variations in the results of participants were Fi g. 5 Results of the intercomparison for 137 Cs; comparison by the ratio of 137 Cs inventory to the dose rate. The horizontal line represents the group mean of each day The error bars represent expanded measurement uncertainty (coverage factor, k = 2) Fi g. 4 Results of the intercomparison for 134 Cs; comparison by the ratio of 134 Cs inventory to the dose rate. The horizontal line represents the group mean of each day. The error bars represent expanded measurement uncertainty (coverage factor, k = 2). Table 6 Percent differences from the group mean for 134 Cs and 137 Cs from the intercomparison on June 10, Participant Code G H ( 134 Cs/ dose rate) ( 137 Cs/ dose rate)

6 In Situ Gamma Spectrometry Intercomparison in Fukushima, Japan 187 Table 7 Results of the intercomparison for natural radionuclides. Statistical 208 Tl 228 Ac 40 K 208 Tl 228 Ac 40 K Parameter 1 ] 1 ] 1 ] 1 ] 1 ] 1 ] Mean 13.5* Std. Devi Coeff. Vari.% Number n Min Max * The branching ratio of Tl is Table 8 Percent differences from the group mean for natural nuclides from the intercomparison on June 3, Participant Code A C D E F 208 Tl Ac K Table 9 Percent differences from the group mean for natural nuclides from the intercomparison on June 10, Participant Code G H 208 Tl Ac K observed for 228 Ac and 208 Tl (Tables 7 and 8). The maximum differences from the group mean were 20% for 208 Tl, 15% for 228 Ac, and 7.0% for 40 K on the rst day (Table 8). Those of the second day were 14% for 208 Tl, 6.0% for 228 Ac, and 1.6% for 40 K (Table 9). It is assumed that the measurements would have been more precise if the measurements with a longer acquisition time were performed, since the count rate of gamma rays from natural radionuclides was less than that from the deposited radiocesium. The results with respect to natural radionuclides were acceptable, although the values of the C. V. were relatively large compared to those of radiocesium, since our main purpose was to determine radiocesium deposition. Furthermore, the results were better than those of a former similar study that show approximately 10% of C. V. for evaluations of 40 K inventory among 24 participants. 8) IV CONCLUSION An intercomparison was performed with eight teams from four institutions which have contributed to the governmentled project to construct distribution maps of radionuclide deposition in the ground. This was accomplished through simultaneous measurement at a speci c site, con rming the measurement accuracy and demonstrating the compatibility among the teams. The inventories evaluated by the participants agreed within 10%, 9%, and 4% for 134 Cs, 137 Cs, and 40 K, respectively, in terms of coef cient of variation without any correction for inhomogeneity. After correction for inhomogeneous distribution of the dose rate in air, the evaluated values agreed within 6% for both 134 Cs and 137 Cs. These intercomparisons among institutions should be continued for the quality assurance of obtained data and con rmation of their consistency. For assessing the degree of agreement in an intercomparison, the uncertainty of the results can be larger due to inhomogeneous distribution of the deposited radionuclides. Therefore, to minimize the uncertainty originating at a site, even with inhomogeneous dose rate distribution, intercomparison with sequential measurements at a speci c site would be more effective than simultaneous measurements for assessing the measurement accuracy, granted the conditions of the site do not change during the intercomparison. ACKNOWLEDGEMENTS This work was a part of the 2013 project entitled the establishment of grasp method of long-term effects caused by radioactive materials from the Fukushima Daiichi Nuclear Power Plant accident, which was nanced by the Nuclear Regulation Authority, Japan. REFERENCES 1) Japan Atomic Energy Agency; Report on the second research on distribution of the radioactive materials from the Fukushima Daiichi Nuclear Power Plant accident (March 2013), ( cat01/entry05.html) (Last accessed 28 August 2015) (in Japanese). 2) Japan Atomic Energy Agency; Report on the 2012 project the establishment of grasp method of long-term effects caused by radioactive materials from the Fukushima Daiichi Nuclear Power Plant accident, ( jaea.go.jp/initiatives/cat03/entry05.html) (Last accessed 28 August 2015) (in Japanese). 3) S. MIKAMI, T. MAEYAMA, Y. HOSHIDE, R. SAKAMOTO, S. SATO, N. OKUDA, S. DEMONGEOT, R. GURRIARAN, Y. UWAMINO, H. KATO, M. FUJIWARA, T. SATO, H. TAKEMIYA and K. SAITO; Spatial distributions of radionuclides deposited onto ground soil around the Fukushima Daiichi Nuclear Power Plant and their temporal change until

7 188 Satoshi MIKAMI, Shoji SATO, Yoshifumi HOSHIDE, Ryuichi SAKAMOTO, Naotoshi OKUDA and Kimiaki SAITO December 2012, J. Environ. Radioact., 139, (2015). 4) H. L. BECK, J. DECAMPO and C. GOGOLAK; In situ Ge(Li) and NaI(Tl) Gamma-ray Spectrometry, HASL-258, Health and Safety Laboratory (USAEC), New York (1972). 5) International Commission on Radiation Units and Measure ments; Gamma-Ray Spectrometry in the Environment, ICRU Report 53, Maryland, USA (1994). 6) K. M. MILLER; Field Gamma-ray Spectrometry, Section 3, Vol. 1, HASL-300, 28th Edition, USDOE (1997). 7) K. M. MILLER, P. SHEBELL, M. A. MONETTI, G. A KLEMIC, R. VENKATARAMAN, E. FISHER, D. G. SCOGGINS, S. H. FALLER, B. MOORE, R. T. REIMAN, D. G. KEEFER and B. GILMARTIN; An intercomparison on in situ gamma-ray spectrometers, Radioact. Radiochem., 9 (4) (1998). 8) H. LETTNER, A. ANDRASI, A. K. HUBMER, E. LOVRANICH, F. STEGER and P. ZOMBORI; In situ gamma spectrometry intercomparison exercise in Salzburg, Austria, Nucl. Instrum. Methods Phys. Res., A369, (1996). 9) Ministry of Education, Culture, Sports, Science and Technology; In-situ measurement method with Germanium semiconductor detector, Radiation Measurement Method Series No. 33 (2007) (in Japanese). 10) Japan Atomic Energy Agency; Report on the 2013 project the establishment of grasp method of long-term effects caused by radioactive materials from the Fukushima Daiichi Nuclear Power Plant accident (March 2014), ( (Last accessed 28 August 2015) (in Japanese). 11) N. MATSUDA, S. MIKAMI, S. SHIMOURA, J. TAKAHASHI, M. NAKANO, K. SHIMADA, K. UNO, S. HAGIWARA and K. SAITO; plate over a wide area surrounding the Fukushima Dai-ichi Nuclear Power Plant, Japan, J. Environ. Radioact., 139, (2015). 12) Joint Committee for Guides in Metrology; Evaluation of measurement data Guide to the expression of uncertainty in measurement, GUM 1995 with minor corrections, JCGM 100 (2008). Satoshi MIKAMI Assistant principal engineer of the Japan Atomic Energy Agency. Since December 2011, Mr. MIKAMI has been involved in the national mapping projects implemented after the Fukushima accident, including repeated large-scale environmental monitoring for clarifying the distributions of radionuclide deposition densities and dose rate in air over wide area, and their time-dependent tendencies. From 2011 to 2015 he was a member of the editorial board for the Japanese Journal of Health Physics. mikami.satoshi@jaea.go.jp