Radiation Protection Dosimetry Advance Access published April 17, 2014 Radiation Protection Dosimetry (2014), pp. 1–6
doi:10.1093/rpd/ncu079
INTERNATIONAL INTERCOMPARISON OF MEASURING INSTRUMENTS FOR RADON/THORON GAS AND RADON SHORT-LIVED DAUGHTER PRODUCTS IN THE NRPI PRAGUE K. Jı´lek*, M. Hy´zˇa, L. Kotı´k, J. Thomas and L. Toma´sˇek National Radiation Protection Institute, Prague, the Czech Republic *Corresponding author: [email protected]
INTRODUCTION Many countries conduct a lot of radon/thoron gas measurements and surveys organised by different institutions with a common goal to find healthy risk dwellings (houses, schools, workplace, etc.) from relevant natural exposure standpoint and subsequently to design cheap and effective remedy. Although passive integral detectors are predominantly used for such measurements in the wide-scale range, special continuous monitors are needed to be used within some special diagnostic investigations in houses, as well. With respect to magnitudes of the effective dose conversion factors, attention must be also paid to measurement of both radon and thoron shortlived daughter products. It is known that thoron gas alone is not so dangerous in practice with respect to its small diffusion length in air, despite it appears as a good indicator for the presence of its short-lived daughter products. That’s why one could pay always attention also to measurement of thoron gas in a house. Since one can expect a mixture of radon/thoron gas in any house, it seems reasonable to pay also attention to measurement instruments capable of distinguishing radon and thorn gas in the mixed field. Until now, no international intercomparison of thoron gas in the mixed field with radon gas has been carried out, yet. MATERIALS AND METHODS Participants In total, 14 laboratories from 11 different countries took part in the 2013 National Radiation Protection Institute (NRPI) intercomparison.
They submitted 5 continuous monitors, 21 electronic detectors and 9 passive integral systems including more than one set of detectors based on both SSNTD and electrets for radon gas measurements carried out in the big NRPI radon chamber. Four laboratories also submitted five continuous monitors for the measurement of short-lived radon decay products in the chamber, there. Each of the four laboratories also submitted more than two sets of pairs of a radon/thoron discriminative passive integral systems based on both SSNTD and electrets for measurement in a small radon/ thoron chamber. Based on all the submitted instruments, the following four types of exposures were carried out during the intercomparison: A: exposure of passive integral radon/thoron discriminative detectors in the small radon/thoron chamber. C: exposure of passive integral detectors on pure radon atmosphere in the big radon chamber. D: exposure of continuous monitors or electronic detectors on pure radon atmosphere in the big radon chamber. E: exposure of continuous monitors to equivalent equilibrium radon concentration (EEC) in the big chamber. The list of participants and corresponding submitted exposures is given in Table 1.
Radon chamber facility The big NRPI radon chamber is a 48 m3 type walk—with airlock inside allowing following quantities to be adjusted, held stable, ON-line monitored and recorded:
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
During the 7th European Conference on Protection Against Radon at Home and at Work held in the autumn of 2013 in Prague, the second intercomparison of measuring instruments for radon and its short-lived decay products and the first intercomparison of radon/thoron gas discriminative passive detectors in mix field of radon/thoron were organised by and held at the Natural Radiation Division of the National Radiation Protection Institute (NRPI) in Prague. In total, 14 laboratories from 11 different countries took part in the 2013 NRPI intercomparison. They submitted both continuous monitors for the measurement of radon gas and equivalent equilibrium radon concentration in a big NRPI chamber (48 m3) and sets of passive detectors including radon/thoron discriminative for the measurement of radon gas in the big chamber and thoron gas in a small thoron chamber (150 dm3).
K. JI´LEK ET AL. Table 1. List of participants and submitted exposures. Institution
Country Czech Republic Poland Poland Poland Italy Italy Germany UK Slovenia Serbia Romania Japan Norway Brasil Czech Republic
C
D
E
X X
X X
X
X
X X
X X X
X X X X/R
X X X X R
X X
X
R
R
X—compared value, R—reference value, A—Rn/Th exposure, C—Rn exposure of passive detectors, only, D—Rn exposure of continuous monitors, only, E—EEC exposure of continuous monitors, only.
– – – – –
adjusted in the chamber to be 8 kBq m23, the relative combined standard uncertainty (K ¼ 1)(2) of reference value given by the NRPI should not exceed 5 %. The results from the reference NRPI monitor Alphaguard calibrated against a Czech radium standard(1) via the NRPI reference scintillation cells NY were used as reference values.
radon concentration, EEC, including its unattached fraction ( fp), air exchange rate (ACH), temperature and relative humidity (RH) and generated spherical aerosols.
Radon gas The chamber inner atmosphere can be adjusted and held stable from 100 to 100 kBq m23. The known and stable steady-state radon concentration is reached by means of defined, stable and measurable ACH and known, constant and adjustable radon entry rate into the chamber. The radon entry rate can be varied in dependence on the use of various types of a dry226Ra/222Rn sources type flow pass through, certified also to their radon production. During the entire exposure, radon concentration in the chamber was kept constant on level of 8 kBq m23 and continuously monitored using both the NRPI reference continuous monitor Alphaguard and by means of grab samples of the chamber inner atmosphere into reference NRPI scintillation cells type NY (Atomki, Hungary). The monitor was adjusted on an hourly basis, and chamber inner atmosphere was sampled twice a day. Whilst the monitor Alphaguard was calibrated on the primary atmosphere of the Physikalisch Technische Bundesanstalt (PTB) Braunschweig, Germany, the scintillation cells were traceable to the NRPI primary measurement standard based on pulsing ion chambers(1) using a water source of 226Ra made in the Czech Metrological Institute. With respect to primary calibration, uncertainties of 3 % (1 sigma) for both instruments and their counting statistics for steady-state radon gas concentration
EEC and fp The NRPI big radon chamber accessories enable to create defined and stable chamber inner atmosphere also from EEC, equilibrium factor F and fp point of view. A carnauba wax aerosol generator is used to maintain a higher aerosol concentration. On the other hand, to produce low equilibrium factor F and high values fp, an electrostatic precipitator including fan is used. Temperature and RH in the chamber can be also separately measured, changed and kept stable during the entire exposure. Whilst temperature can be easily adjusted, changed and kept stable in the range of 8 to 45 8C by means of a heating and cooling system, the RH can be changed, measured and kept stable in the range of 5 to 95 % by means of a proper HW and SW using measurement instrument TESTO Hygrotest HG 600. PHT, allowing both measurement of temperature and RH and driving connected humidifier and dehumidifier. During the entire exposure, equilibrium factor F and fp were almost kept constant by means of injection of aerosols from the chamber built in the carnauba wax aerosol generator. About 20 h before the end of the entire exposure, the generator was switched OFF. Temperature and RH were not be kept stable during the exposure and temperature ranged 27 8C and RH
Page 2 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Nuclear Physics Institute of the CAS Institute of Nuclear Physics PAN Central Laboratory for Radiation Protection The Nofer Institute of Occupational Medicine MI.AM Srl Agenzia Regionale per la Protezione del Ambiente FVG Helmoltz Centrum Mu¨nchen Track Analysis Systems Ltd. ZVD Zavod za Varstvo pri Delu D.D. Faculty of Sciences, Univ. of Novi Sad University Babes-Bolyai National Institute of Radiological Science, Chiba Corentium a.s. Laboratory of Poqos de Caldas, CNEN National Radiation Protection Institute of Prague
A
RADON PROGENY SIZE DISTRIBUTIONS AND ENHANCED DEPOSITION EFFECTS
The chamber shell has built-in one multi-pin vacuum and three gas inlet/outlet bushings. The vacuum bushing allows the use of HV power supply in the vessel and monitoring of both RH and temperature and radon/thoron gas concentration by means of a proper monitors and sensors deployed inside the vessel. The three inlet/outlet bushings with closing taps allow external chamber filling and samplings. Inside the vessel, a stainless steel shelf can be also deployed allowing to create three floors that can be independently shifted up or down along the vertical axe. The chamber is totally free of any gas leakage. Homogeneity Homogeneity for both radon and thoron gas inside the chamber was many times checked using a large number of exposed radon/thoron discriminative electret ion chambers recently developed at the NRPI (see Figure 1) and statistically confirmed adopting the quasi-Poisson regression model for data with observed overdispersion. The mentioned radon/thoron gas discriminative electret ion chamber system is in fact modification of a Czech commercially available radon gas integral electret system RM-1TM having been used by the NRPI for more than 15 y also in the scope of the current National Radon Programme. Generally, radon/thoron gas homogeneity is assured inside the chamber with proper position of both used emanation Rn/Tn source and exposed detectors on the floors of the shelf inside the chamber and further with the use of three fans deployed inside the chamber properly along its vertical profile. Radon/thoron gas During the entire exposure, very stable mixed field of radon/thoron gas concentration was created inside
Thoron chamber facility The thoron NRPI chamber comprises 150-dm3 stainless steel cylindrical vessel of 80 cm in height and 45 cm in diameter without external thermal isolation. The chamber has been established in the NRPI as a small radon chamber several years ago(1) and was frequently used also for the calibration of radon passive radon integral detectors.
Figure 1. Discriminative radon/thorn gas electret ion chamber system developed at the NRPI.
Page 3 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
of 45 %. During the exercise, all tested instruments were placed on a table at the centre of the chamber. The EEC, F and fp were continuously ON-line monitored during the whole exposure by means of the NRPI reference continuous monitor Fritra 4. The monitor allows measurement of radon gas and unattached and attached activity of each short-lived decay product. The monitor alone was calibrated against the NRPI reference radon daughter product measurement instrument RRDPMI(1, 3) both for EEC and fp and also independently successfully compared with primary standards of renowned laboratories as the PTB, BfS Berlin, Germany and the Authorized Metrological Centre (AMC-Kamenna´, Czech Republic). Having in mind primary calibration uncertainty of EEC better than 8 % (1 sigma) for monitor Fritra 4 and its counting statistics better than 4 % for measured level of EEC of 4 kBq m23, the combined relative standard uncertainty (K ¼ 1) of the NRPI reference values EEC should not exceed 10 %. Besides on-line EEC and fp monitoring of the chamber inner atmosphere, also twice a day one-grab samplings were taken from the chamber through the diffusion screen on the Millipore AA, 0.8-mm filter placed behind the screen to estimate unattached and attached activities of each short-lived radon decay product including fp. Additionally, total aerosol concentration and aerosol size distribution were monitored by means of Scanning Perticle Mobility Sizer SPMS þC (Grimm D) allowing sorted aerosols ranging from 10 to 1100 nm according to their mobility diameters into 44 channels. An independent control of measured results of F, EEC and fp in the chamber is also assured applying of numerical method developed in the NRPI based on the algebraic inversion of the Jacobi-Porstendo¨rfer room model(4, 5). As input data for calculations, the authors used their results of measured attached and unattached activities of each short-lived daughter product and values of ACH measured by means of the tracer gas method(6). Another tool to check the results of F, fp provides also results of the measurement aerosol size distribution and use of the method based on aerosol attached theory proposed by Porstendo¨rfer(7). The chamber description in more details with all its accessories has been already described elsewhere(3).
K. JI´LEK ET AL. 226
228
chamber inner average temperature ranged during all the exposure 258C, the average RH was 6 % due to externally connected RAD7 with its drier tube. With respect to expected stable and known radon/ thoron gas entry rates into the chamber, investigated average values of radon/thoron gas concentrations could be also calculated independently on their direct measurements. The NRPI reference monitor RAD7 alone has been recently calibrated on the primary thoron gas atmosphere at the PTB Braunschweig. Additionally, the current NRPI radon/thoron gas facilities also allow carrying out an independent RAD 7 thoron gas calibration, at any time. Independent calibrations are principally based on the use of both proper vessels including also the small radon/thoron chamber, precision flow meter calibrator Defender model 530 BIOS (Mesa Lab. USA) and certified pure radon and thoron gas sources type flow pass through produced by either the Czech Metrological Institute or by the Pylon (Canada). Additionally, the NRPI has been recently also equipped with a pure thoron gas emanation source produced in the PTB and based on 228Th matrix. The source has 228Th activity of 50 kBq and emanation coefficient of 40 % estimated again by means of the above-mentioned NRPI HPGE gamma spectroscopy with total uncertainty budged of measurement of ,3 % for each measurement quantity. Overall the uncertainty of each separately measured average values of radon/thoron gas concentration represented by relative combined standard uncertainty for K ¼ 1, the NRPI declares better than 10 %. All the exposure conditions and magnitudes both in the big radon chamber and the small radon/thoron chamber are summarised in Table 2.
Table 2. Exposition conditions and magnitudes. Exposure type
A/Rn
A/Th
Exposure duration (h) av/EEC (kBq m23) SD (kBq m23) SEM (kBq m23) ACH (h21) RH (%) T (8C) fp F Z (cm23) aerosols GM (nm)/GSD
25 1.34 0.88 0.05
25 25.5 2.2 0.1
6 26.8 .0.65 ,0.05 ,600
6 26.8 .0.65 ,0.05 ,600
C
D
E
67.5 8.22 0.39 0.05 0.07 44.1 27.4 0.06 0.05 6750 125/1.8
72.5 8.30 0.48 0.06 0.07 44.3 27.4 0.06 0.05 6670 138/1.9
72.5 4.30 0.79 0.09 0.07 44.3 27.4 0.06 0.05 6670 138/1.9
av/EEC—mean values of radon and equivalent equilibrium radon gas concentration, respectively; SD, SEM—relevant standard deviation and standard error of the mean, respectively; ACH—mean value of air exchange rate; RH,T—mean values of RH and temperature, respectively; fp,F—mean values of unattached fraction of EEC and equilibrium factor, respectively; Z—total aerosol concentration, GM/GSD—parameters of aerosol size distribution.
Page 4 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
the chamber by means of used mix Ra/ Th emanation source produced as prototype for the NRPI at the Czech Metrological Institute. The source alone allows creating thoron gas concentration of the order 10 kBq m23 and radon gas concentration of the order 1 kBq m23 inside the chamber within 1 d. Both activity of 226Ra and 228Th, respectively, and relevant emanation coefficients for radon and thoron gas were estimated by means of the HPGE gamma spectroscopy with high resolution for 60Co of 1332 keV better than 1.8 keV. The emanation coefficients were calculated from activity ratios of 224Ra/212Pb and 226Ra/214Pb, respectively, obtained by analysing proper gamma peaks. The total uncertainty budged of measurement was estimated to be ,3 % in both cases. The results of measurement of the emanation coefficients in ambient conditions similar to those used in the chamber during the intercomparison indicated for radon gas mean value exceeded 99 %, and for thoron gas value exceeded 75 %. After deployment of all the sets of paired discriminative radon/thoron passive integral detectors on proper floors of the chamber shelf, the emanation source gas was immediately placed into the chamber. When the chamber was closed, the reference NRPI radon/thoron gas continuous monitor RAD7 was connected to the air inlet/outlet bushings by means of short loop and switched ON. During the entire exposure, the chamber inner atmosphere was continuously monitored with the RAD 7 set-up in the sniffing mode THORON allowing 5-min records. At the same time, both temperature and RH were recorded on hourly basis by means of the chamber built-in sensors, as well. Whilst the
RADON PROGENY SIZE DISTRIBUTIONS AND ENHANCED DEPOSITION EFFECTS Table 3. Median values of R, var R, ERRb and zeta score and their range in parenthesis. Exposure type A
Rn Th Rn Rn EEC
C D E
R
var R
ERRb (%)
z—zeta score
1.06 (0.98–1.73) 0.95 (0.91–1.08) 1.02 (0.84–1.24) 1.02 (0.92–1.10) 1.01(0.81–1.06)
0.15 (0.08–0.35) 0.05 (0.04–0.19) 0.03 (0.02–0.26) 0.02 (0.02–0.04) 0.06 (0.04–0.06)
6 (22 to 73) 25 (29 to 8) 2 (216 to 24) 2 (28 to 11) 1 (219 to 6)
0.7 (20.5 to 4.4) 22.0 (23.8 to 0.8) 1.2 (211.9 to 19.1) 1.4 (29.8 to 10.5) 0.3 (27.6 to 2.0)
R, var R—ratio of the means and corresponding variance R, respectively; ERRb—biased error ; A,C,D,E—exposures with the same meaning as in Table 2.
After the end of the entire exposures, all the exposed monitors and detectors including transit detectors were returned to the originating laboratories for the evaluation of results in terms of relevant mean values, corresponding standard deviations (SDs) and standard errors of the mean (SEM). Whilst SDs and SEM were calculated from the results of exposed sets of passive integral dosemeters, the SD and SEM of the used continuous monitor were calculated from all the recorded values adjusted on an hourly basis in most cases.
RESULTS To evaluate participants’ data and compare them with the NRPI reference values by means of either mean values or within the uncertainty claimed by each participant, the authors define mean deviation R, its variance R (var R), biased error (ERRb) (%) and zeta score (z)(8) as follows: R ¼ X /Y
ð1Þ
ERRb ¼ ðR 1Þ 100 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X2 Sx 2 Sy 2 pffiffiffi þ pffiffiffiffi var R ¼ 2 Y X n Y m
ð2Þ
X Y ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 6 ¼ s 2 Sy Sx 2 pffiffiffi þ pffiffiffiffi X n Y m
ð3Þ
Figure 2. Proportions of institutions by achieved rank for each exposure and mean R.
Fieller’s theorem(9, 10), in a usual way as R + 2sR, where sR is the estimated SD of R, i.e. the square root of the variance R (var R) estimate in Equation (3). The practical interpretation of the absolute value of the zeta score [abs(z)] is as follows: – – –
abs(z) 3 indicates unsatisfactory results, abs(z) 2 indicates satisfactory results and 2 abs(z) 3 indicates questionable results.
ð4Þ
where X and Y are mean values reported by participants and the NRPI is reference mean value, respectively. sX and sY are sample SDs from a given institute and from NRPI, respectively, and n and m are the corresponding sample sizes. The approximate 95 % confidence interval for ratio of the two means (R) is constructed adopting the
Median values of R, var R, ERRb and abs(z) and their range for corresponding exposures are given in Table 3. Based on both practically ranked values of R defined below and the zeta score, the authors illustrated in Figures 2 and 3 proportions of institutions by achieved relevant rank for each exposure. R5% 5 % , R 10 % 10 % , R 20 % R . 20 %
Page 5 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Logistical arrangement
K. JI´LEK ET AL.
average agreement (average deviation up to 6 %) with the NRPI reference value. Only one monitor was out of acceptable 15 % average agreement, and it needs to be recalibrated. All the above-mentioned conclusions were also supported with corresponding values of zeta score. The medians of absolute values of the zeta scores calculated for all the selected types of conducted exposures were found to be ,2. FUNDING
REFERENCES
CONCLUSIONS Based on the results given in Table 3, the following can be concluded: Radon/thoron intercomparison Exposure A: the sets of radon/thoron discriminative detectors belonging to three of four participating institutions achieved a very good acceptable agreement (average deviation of up to 15 %) with the NRPI reference value both for radon and for thoron gas. Only one institution had problems with results, and its radon gas results were far from the reference value. Radon gas intercomparison Exposure C: the results of .50 % of the all exposed sets of passive detectors indicated very well average agreement (average deviation up to 5 %) with the NRPI reference value. The results of the rest of institutions ranged up to acceptable average deviation of 15 % from the NRPI reference value. Only results of one of nine institutions indicated .20 % systematic overestimation from the reference value for the both types of used SSNTD. Exposure D: the results of .50 % of nine exposed radon continuous monitors indicated very well average agreement (average deviation up to 5 %) with the NRPI reference value. The results of the rest of the institutions ranged up to an acceptable average deviation of 10 %. Exposure E: the results of four from five exposed continuous monitors’ EEC indicated a very good
1. Jilek, K., Thomas, J. and Brabec, M. Quality assurance programme for radon and its short-lived progeny measuring instruments in NRPI Prague. Rad. Prot. Dosim. 130(1), 43–47 (2008). 2. ISO Evaluation of measurement data. Guide to the expression of uncertainty in measurements (GUM), JCGM 100: 2008 (ISO/IEC Guide 98–3) (2008). 3. Jilek, K. and Marusiakova, M. Results of the 2010 National Radiation Protection Institute intercomparison of radon and its short-lived decay product continuous monitors. Radiat. Prot. Dosim. 145(2–3), 273–279 (2011). 4. Thomas, J. and Jilek, K. Evaluation and comparison of measurements of unattached and attached radon progeny in the radon chamber of PTB Braunschweig (Germany) with NRPI Praha (Czech Republic). Radiat. Prot. Dosim. 145(2– 3), 316– 319 (2011). 5. Thomas, J. and Jilek, K. Invariants of the JacobiPorstendo¨rfer room model for radon progeny in indoor air. Radiat. Prot. Dosim. 150(2), 142–149 (2012). 6. Brabec, M. and Jilek, K. State-space dynamic model for estimation of radon entry rate, based on Kalman filtering. J. Environ. Radioact. 98, 285–297 (2007). 7. Porstendo¨rfer, J. Physical parameters and dose factors of the radon and thoron decay products. Radiat. Prot. Dosim. 94(4), 365–373 (2001). 8. ISO 13528:2003. Statistical methods for use in proficiency testing by interlaboratory comparisons. ISO 2003. Available at URL: http://www.bipm.org/utils/ common/documents/jcgm. 9. Fieller, E. C. The distribution of the index in a bivariate Normal distribution. Biometrika 24(3– 4), 428–440 (1932). 10. Fieller, E. C. A fundamental formula in the statistics of biological assay, and some applications. Quart. J. Pharm. Pharmacol. 17, 117–123 (1944).
Page 6 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Figure 3. Proportions of institutions by achieved rank for each exposure and z—zeta score.
The present work has been partially funded by the Grant from the Technological Agency of the Czech ˇ R) under the Contract No. TA Republic (TAC 02020865.
doi:10.1093/rpd/ncu079
INTERNATIONAL INTERCOMPARISON OF MEASURING INSTRUMENTS FOR RADON/THORON GAS AND RADON SHORT-LIVED DAUGHTER PRODUCTS IN THE NRPI PRAGUE K. Jı´lek*, M. Hy´zˇa, L. Kotı´k, J. Thomas and L. Toma´sˇek National Radiation Protection Institute, Prague, the Czech Republic *Corresponding author: [email protected]
INTRODUCTION Many countries conduct a lot of radon/thoron gas measurements and surveys organised by different institutions with a common goal to find healthy risk dwellings (houses, schools, workplace, etc.) from relevant natural exposure standpoint and subsequently to design cheap and effective remedy. Although passive integral detectors are predominantly used for such measurements in the wide-scale range, special continuous monitors are needed to be used within some special diagnostic investigations in houses, as well. With respect to magnitudes of the effective dose conversion factors, attention must be also paid to measurement of both radon and thoron shortlived daughter products. It is known that thoron gas alone is not so dangerous in practice with respect to its small diffusion length in air, despite it appears as a good indicator for the presence of its short-lived daughter products. That’s why one could pay always attention also to measurement of thoron gas in a house. Since one can expect a mixture of radon/thoron gas in any house, it seems reasonable to pay also attention to measurement instruments capable of distinguishing radon and thorn gas in the mixed field. Until now, no international intercomparison of thoron gas in the mixed field with radon gas has been carried out, yet. MATERIALS AND METHODS Participants In total, 14 laboratories from 11 different countries took part in the 2013 National Radiation Protection Institute (NRPI) intercomparison.
They submitted 5 continuous monitors, 21 electronic detectors and 9 passive integral systems including more than one set of detectors based on both SSNTD and electrets for radon gas measurements carried out in the big NRPI radon chamber. Four laboratories also submitted five continuous monitors for the measurement of short-lived radon decay products in the chamber, there. Each of the four laboratories also submitted more than two sets of pairs of a radon/thoron discriminative passive integral systems based on both SSNTD and electrets for measurement in a small radon/ thoron chamber. Based on all the submitted instruments, the following four types of exposures were carried out during the intercomparison: A: exposure of passive integral radon/thoron discriminative detectors in the small radon/thoron chamber. C: exposure of passive integral detectors on pure radon atmosphere in the big radon chamber. D: exposure of continuous monitors or electronic detectors on pure radon atmosphere in the big radon chamber. E: exposure of continuous monitors to equivalent equilibrium radon concentration (EEC) in the big chamber. The list of participants and corresponding submitted exposures is given in Table 1.
Radon chamber facility The big NRPI radon chamber is a 48 m3 type walk—with airlock inside allowing following quantities to be adjusted, held stable, ON-line monitored and recorded:
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
During the 7th European Conference on Protection Against Radon at Home and at Work held in the autumn of 2013 in Prague, the second intercomparison of measuring instruments for radon and its short-lived decay products and the first intercomparison of radon/thoron gas discriminative passive detectors in mix field of radon/thoron were organised by and held at the Natural Radiation Division of the National Radiation Protection Institute (NRPI) in Prague. In total, 14 laboratories from 11 different countries took part in the 2013 NRPI intercomparison. They submitted both continuous monitors for the measurement of radon gas and equivalent equilibrium radon concentration in a big NRPI chamber (48 m3) and sets of passive detectors including radon/thoron discriminative for the measurement of radon gas in the big chamber and thoron gas in a small thoron chamber (150 dm3).
K. JI´LEK ET AL. Table 1. List of participants and submitted exposures. Institution
Country Czech Republic Poland Poland Poland Italy Italy Germany UK Slovenia Serbia Romania Japan Norway Brasil Czech Republic
C
D
E
X X
X X
X
X
X X
X X X
X X X X/R
X X X X R
X X
X
R
R
X—compared value, R—reference value, A—Rn/Th exposure, C—Rn exposure of passive detectors, only, D—Rn exposure of continuous monitors, only, E—EEC exposure of continuous monitors, only.
– – – – –
adjusted in the chamber to be 8 kBq m23, the relative combined standard uncertainty (K ¼ 1)(2) of reference value given by the NRPI should not exceed 5 %. The results from the reference NRPI monitor Alphaguard calibrated against a Czech radium standard(1) via the NRPI reference scintillation cells NY were used as reference values.
radon concentration, EEC, including its unattached fraction ( fp), air exchange rate (ACH), temperature and relative humidity (RH) and generated spherical aerosols.
Radon gas The chamber inner atmosphere can be adjusted and held stable from 100 to 100 kBq m23. The known and stable steady-state radon concentration is reached by means of defined, stable and measurable ACH and known, constant and adjustable radon entry rate into the chamber. The radon entry rate can be varied in dependence on the use of various types of a dry226Ra/222Rn sources type flow pass through, certified also to their radon production. During the entire exposure, radon concentration in the chamber was kept constant on level of 8 kBq m23 and continuously monitored using both the NRPI reference continuous monitor Alphaguard and by means of grab samples of the chamber inner atmosphere into reference NRPI scintillation cells type NY (Atomki, Hungary). The monitor was adjusted on an hourly basis, and chamber inner atmosphere was sampled twice a day. Whilst the monitor Alphaguard was calibrated on the primary atmosphere of the Physikalisch Technische Bundesanstalt (PTB) Braunschweig, Germany, the scintillation cells were traceable to the NRPI primary measurement standard based on pulsing ion chambers(1) using a water source of 226Ra made in the Czech Metrological Institute. With respect to primary calibration, uncertainties of 3 % (1 sigma) for both instruments and their counting statistics for steady-state radon gas concentration
EEC and fp The NRPI big radon chamber accessories enable to create defined and stable chamber inner atmosphere also from EEC, equilibrium factor F and fp point of view. A carnauba wax aerosol generator is used to maintain a higher aerosol concentration. On the other hand, to produce low equilibrium factor F and high values fp, an electrostatic precipitator including fan is used. Temperature and RH in the chamber can be also separately measured, changed and kept stable during the entire exposure. Whilst temperature can be easily adjusted, changed and kept stable in the range of 8 to 45 8C by means of a heating and cooling system, the RH can be changed, measured and kept stable in the range of 5 to 95 % by means of a proper HW and SW using measurement instrument TESTO Hygrotest HG 600. PHT, allowing both measurement of temperature and RH and driving connected humidifier and dehumidifier. During the entire exposure, equilibrium factor F and fp were almost kept constant by means of injection of aerosols from the chamber built in the carnauba wax aerosol generator. About 20 h before the end of the entire exposure, the generator was switched OFF. Temperature and RH were not be kept stable during the exposure and temperature ranged 27 8C and RH
Page 2 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Nuclear Physics Institute of the CAS Institute of Nuclear Physics PAN Central Laboratory for Radiation Protection The Nofer Institute of Occupational Medicine MI.AM Srl Agenzia Regionale per la Protezione del Ambiente FVG Helmoltz Centrum Mu¨nchen Track Analysis Systems Ltd. ZVD Zavod za Varstvo pri Delu D.D. Faculty of Sciences, Univ. of Novi Sad University Babes-Bolyai National Institute of Radiological Science, Chiba Corentium a.s. Laboratory of Poqos de Caldas, CNEN National Radiation Protection Institute of Prague
A
RADON PROGENY SIZE DISTRIBUTIONS AND ENHANCED DEPOSITION EFFECTS
The chamber shell has built-in one multi-pin vacuum and three gas inlet/outlet bushings. The vacuum bushing allows the use of HV power supply in the vessel and monitoring of both RH and temperature and radon/thoron gas concentration by means of a proper monitors and sensors deployed inside the vessel. The three inlet/outlet bushings with closing taps allow external chamber filling and samplings. Inside the vessel, a stainless steel shelf can be also deployed allowing to create three floors that can be independently shifted up or down along the vertical axe. The chamber is totally free of any gas leakage. Homogeneity Homogeneity for both radon and thoron gas inside the chamber was many times checked using a large number of exposed radon/thoron discriminative electret ion chambers recently developed at the NRPI (see Figure 1) and statistically confirmed adopting the quasi-Poisson regression model for data with observed overdispersion. The mentioned radon/thoron gas discriminative electret ion chamber system is in fact modification of a Czech commercially available radon gas integral electret system RM-1TM having been used by the NRPI for more than 15 y also in the scope of the current National Radon Programme. Generally, radon/thoron gas homogeneity is assured inside the chamber with proper position of both used emanation Rn/Tn source and exposed detectors on the floors of the shelf inside the chamber and further with the use of three fans deployed inside the chamber properly along its vertical profile. Radon/thoron gas During the entire exposure, very stable mixed field of radon/thoron gas concentration was created inside
Thoron chamber facility The thoron NRPI chamber comprises 150-dm3 stainless steel cylindrical vessel of 80 cm in height and 45 cm in diameter without external thermal isolation. The chamber has been established in the NRPI as a small radon chamber several years ago(1) and was frequently used also for the calibration of radon passive radon integral detectors.
Figure 1. Discriminative radon/thorn gas electret ion chamber system developed at the NRPI.
Page 3 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
of 45 %. During the exercise, all tested instruments were placed on a table at the centre of the chamber. The EEC, F and fp were continuously ON-line monitored during the whole exposure by means of the NRPI reference continuous monitor Fritra 4. The monitor allows measurement of radon gas and unattached and attached activity of each short-lived decay product. The monitor alone was calibrated against the NRPI reference radon daughter product measurement instrument RRDPMI(1, 3) both for EEC and fp and also independently successfully compared with primary standards of renowned laboratories as the PTB, BfS Berlin, Germany and the Authorized Metrological Centre (AMC-Kamenna´, Czech Republic). Having in mind primary calibration uncertainty of EEC better than 8 % (1 sigma) for monitor Fritra 4 and its counting statistics better than 4 % for measured level of EEC of 4 kBq m23, the combined relative standard uncertainty (K ¼ 1) of the NRPI reference values EEC should not exceed 10 %. Besides on-line EEC and fp monitoring of the chamber inner atmosphere, also twice a day one-grab samplings were taken from the chamber through the diffusion screen on the Millipore AA, 0.8-mm filter placed behind the screen to estimate unattached and attached activities of each short-lived radon decay product including fp. Additionally, total aerosol concentration and aerosol size distribution were monitored by means of Scanning Perticle Mobility Sizer SPMS þC (Grimm D) allowing sorted aerosols ranging from 10 to 1100 nm according to their mobility diameters into 44 channels. An independent control of measured results of F, EEC and fp in the chamber is also assured applying of numerical method developed in the NRPI based on the algebraic inversion of the Jacobi-Porstendo¨rfer room model(4, 5). As input data for calculations, the authors used their results of measured attached and unattached activities of each short-lived daughter product and values of ACH measured by means of the tracer gas method(6). Another tool to check the results of F, fp provides also results of the measurement aerosol size distribution and use of the method based on aerosol attached theory proposed by Porstendo¨rfer(7). The chamber description in more details with all its accessories has been already described elsewhere(3).
K. JI´LEK ET AL. 226
228
chamber inner average temperature ranged during all the exposure 258C, the average RH was 6 % due to externally connected RAD7 with its drier tube. With respect to expected stable and known radon/ thoron gas entry rates into the chamber, investigated average values of radon/thoron gas concentrations could be also calculated independently on their direct measurements. The NRPI reference monitor RAD7 alone has been recently calibrated on the primary thoron gas atmosphere at the PTB Braunschweig. Additionally, the current NRPI radon/thoron gas facilities also allow carrying out an independent RAD 7 thoron gas calibration, at any time. Independent calibrations are principally based on the use of both proper vessels including also the small radon/thoron chamber, precision flow meter calibrator Defender model 530 BIOS (Mesa Lab. USA) and certified pure radon and thoron gas sources type flow pass through produced by either the Czech Metrological Institute or by the Pylon (Canada). Additionally, the NRPI has been recently also equipped with a pure thoron gas emanation source produced in the PTB and based on 228Th matrix. The source has 228Th activity of 50 kBq and emanation coefficient of 40 % estimated again by means of the above-mentioned NRPI HPGE gamma spectroscopy with total uncertainty budged of measurement of ,3 % for each measurement quantity. Overall the uncertainty of each separately measured average values of radon/thoron gas concentration represented by relative combined standard uncertainty for K ¼ 1, the NRPI declares better than 10 %. All the exposure conditions and magnitudes both in the big radon chamber and the small radon/thoron chamber are summarised in Table 2.
Table 2. Exposition conditions and magnitudes. Exposure type
A/Rn
A/Th
Exposure duration (h) av/EEC (kBq m23) SD (kBq m23) SEM (kBq m23) ACH (h21) RH (%) T (8C) fp F Z (cm23) aerosols GM (nm)/GSD
25 1.34 0.88 0.05
25 25.5 2.2 0.1
6 26.8 .0.65 ,0.05 ,600
6 26.8 .0.65 ,0.05 ,600
C
D
E
67.5 8.22 0.39 0.05 0.07 44.1 27.4 0.06 0.05 6750 125/1.8
72.5 8.30 0.48 0.06 0.07 44.3 27.4 0.06 0.05 6670 138/1.9
72.5 4.30 0.79 0.09 0.07 44.3 27.4 0.06 0.05 6670 138/1.9
av/EEC—mean values of radon and equivalent equilibrium radon gas concentration, respectively; SD, SEM—relevant standard deviation and standard error of the mean, respectively; ACH—mean value of air exchange rate; RH,T—mean values of RH and temperature, respectively; fp,F—mean values of unattached fraction of EEC and equilibrium factor, respectively; Z—total aerosol concentration, GM/GSD—parameters of aerosol size distribution.
Page 4 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
the chamber by means of used mix Ra/ Th emanation source produced as prototype for the NRPI at the Czech Metrological Institute. The source alone allows creating thoron gas concentration of the order 10 kBq m23 and radon gas concentration of the order 1 kBq m23 inside the chamber within 1 d. Both activity of 226Ra and 228Th, respectively, and relevant emanation coefficients for radon and thoron gas were estimated by means of the HPGE gamma spectroscopy with high resolution for 60Co of 1332 keV better than 1.8 keV. The emanation coefficients were calculated from activity ratios of 224Ra/212Pb and 226Ra/214Pb, respectively, obtained by analysing proper gamma peaks. The total uncertainty budged of measurement was estimated to be ,3 % in both cases. The results of measurement of the emanation coefficients in ambient conditions similar to those used in the chamber during the intercomparison indicated for radon gas mean value exceeded 99 %, and for thoron gas value exceeded 75 %. After deployment of all the sets of paired discriminative radon/thoron passive integral detectors on proper floors of the chamber shelf, the emanation source gas was immediately placed into the chamber. When the chamber was closed, the reference NRPI radon/thoron gas continuous monitor RAD7 was connected to the air inlet/outlet bushings by means of short loop and switched ON. During the entire exposure, the chamber inner atmosphere was continuously monitored with the RAD 7 set-up in the sniffing mode THORON allowing 5-min records. At the same time, both temperature and RH were recorded on hourly basis by means of the chamber built-in sensors, as well. Whilst the
RADON PROGENY SIZE DISTRIBUTIONS AND ENHANCED DEPOSITION EFFECTS Table 3. Median values of R, var R, ERRb and zeta score and their range in parenthesis. Exposure type A
Rn Th Rn Rn EEC
C D E
R
var R
ERRb (%)
z—zeta score
1.06 (0.98–1.73) 0.95 (0.91–1.08) 1.02 (0.84–1.24) 1.02 (0.92–1.10) 1.01(0.81–1.06)
0.15 (0.08–0.35) 0.05 (0.04–0.19) 0.03 (0.02–0.26) 0.02 (0.02–0.04) 0.06 (0.04–0.06)
6 (22 to 73) 25 (29 to 8) 2 (216 to 24) 2 (28 to 11) 1 (219 to 6)
0.7 (20.5 to 4.4) 22.0 (23.8 to 0.8) 1.2 (211.9 to 19.1) 1.4 (29.8 to 10.5) 0.3 (27.6 to 2.0)
R, var R—ratio of the means and corresponding variance R, respectively; ERRb—biased error ; A,C,D,E—exposures with the same meaning as in Table 2.
After the end of the entire exposures, all the exposed monitors and detectors including transit detectors were returned to the originating laboratories for the evaluation of results in terms of relevant mean values, corresponding standard deviations (SDs) and standard errors of the mean (SEM). Whilst SDs and SEM were calculated from the results of exposed sets of passive integral dosemeters, the SD and SEM of the used continuous monitor were calculated from all the recorded values adjusted on an hourly basis in most cases.
RESULTS To evaluate participants’ data and compare them with the NRPI reference values by means of either mean values or within the uncertainty claimed by each participant, the authors define mean deviation R, its variance R (var R), biased error (ERRb) (%) and zeta score (z)(8) as follows: R ¼ X /Y
ð1Þ
ERRb ¼ ðR 1Þ 100 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi X2 Sx 2 Sy 2 pffiffiffi þ pffiffiffiffi var R ¼ 2 Y X n Y m
ð2Þ
X Y ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 6 ¼ s 2 Sy Sx 2 pffiffiffi þ pffiffiffiffi X n Y m
ð3Þ
Figure 2. Proportions of institutions by achieved rank for each exposure and mean R.
Fieller’s theorem(9, 10), in a usual way as R + 2sR, where sR is the estimated SD of R, i.e. the square root of the variance R (var R) estimate in Equation (3). The practical interpretation of the absolute value of the zeta score [abs(z)] is as follows: – – –
abs(z) 3 indicates unsatisfactory results, abs(z) 2 indicates satisfactory results and 2 abs(z) 3 indicates questionable results.
ð4Þ
where X and Y are mean values reported by participants and the NRPI is reference mean value, respectively. sX and sY are sample SDs from a given institute and from NRPI, respectively, and n and m are the corresponding sample sizes. The approximate 95 % confidence interval for ratio of the two means (R) is constructed adopting the
Median values of R, var R, ERRb and abs(z) and their range for corresponding exposures are given in Table 3. Based on both practically ranked values of R defined below and the zeta score, the authors illustrated in Figures 2 and 3 proportions of institutions by achieved relevant rank for each exposure. R5% 5 % , R 10 % 10 % , R 20 % R . 20 %
Page 5 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Logistical arrangement
K. JI´LEK ET AL.
average agreement (average deviation up to 6 %) with the NRPI reference value. Only one monitor was out of acceptable 15 % average agreement, and it needs to be recalibrated. All the above-mentioned conclusions were also supported with corresponding values of zeta score. The medians of absolute values of the zeta scores calculated for all the selected types of conducted exposures were found to be ,2. FUNDING
REFERENCES
CONCLUSIONS Based on the results given in Table 3, the following can be concluded: Radon/thoron intercomparison Exposure A: the sets of radon/thoron discriminative detectors belonging to three of four participating institutions achieved a very good acceptable agreement (average deviation of up to 15 %) with the NRPI reference value both for radon and for thoron gas. Only one institution had problems with results, and its radon gas results were far from the reference value. Radon gas intercomparison Exposure C: the results of .50 % of the all exposed sets of passive detectors indicated very well average agreement (average deviation up to 5 %) with the NRPI reference value. The results of the rest of institutions ranged up to acceptable average deviation of 15 % from the NRPI reference value. Only results of one of nine institutions indicated .20 % systematic overestimation from the reference value for the both types of used SSNTD. Exposure D: the results of .50 % of nine exposed radon continuous monitors indicated very well average agreement (average deviation up to 5 %) with the NRPI reference value. The results of the rest of the institutions ranged up to an acceptable average deviation of 10 %. Exposure E: the results of four from five exposed continuous monitors’ EEC indicated a very good
1. Jilek, K., Thomas, J. and Brabec, M. Quality assurance programme for radon and its short-lived progeny measuring instruments in NRPI Prague. Rad. Prot. Dosim. 130(1), 43–47 (2008). 2. ISO Evaluation of measurement data. Guide to the expression of uncertainty in measurements (GUM), JCGM 100: 2008 (ISO/IEC Guide 98–3) (2008). 3. Jilek, K. and Marusiakova, M. Results of the 2010 National Radiation Protection Institute intercomparison of radon and its short-lived decay product continuous monitors. Radiat. Prot. Dosim. 145(2–3), 273–279 (2011). 4. Thomas, J. and Jilek, K. Evaluation and comparison of measurements of unattached and attached radon progeny in the radon chamber of PTB Braunschweig (Germany) with NRPI Praha (Czech Republic). Radiat. Prot. Dosim. 145(2– 3), 316– 319 (2011). 5. Thomas, J. and Jilek, K. Invariants of the JacobiPorstendo¨rfer room model for radon progeny in indoor air. Radiat. Prot. Dosim. 150(2), 142–149 (2012). 6. Brabec, M. and Jilek, K. State-space dynamic model for estimation of radon entry rate, based on Kalman filtering. J. Environ. Radioact. 98, 285–297 (2007). 7. Porstendo¨rfer, J. Physical parameters and dose factors of the radon and thoron decay products. Radiat. Prot. Dosim. 94(4), 365–373 (2001). 8. ISO 13528:2003. Statistical methods for use in proficiency testing by interlaboratory comparisons. ISO 2003. Available at URL: http://www.bipm.org/utils/ common/documents/jcgm. 9. Fieller, E. C. The distribution of the index in a bivariate Normal distribution. Biometrika 24(3– 4), 428–440 (1932). 10. Fieller, E. C. A fundamental formula in the statistics of biological assay, and some applications. Quart. J. Pharm. Pharmacol. 17, 117–123 (1944).
Page 6 of 6
Downloaded from http://rpd.oxfordjournals.org/ at Anadolu University on May 14, 2014
Figure 3. Proportions of institutions by achieved rank for each exposure and z—zeta score.
The present work has been partially funded by the Grant from the Technological Agency of the Czech ˇ R) under the Contract No. TA Republic (TAC 02020865.
