Skip to main content
PMC home page
Journal of Radiation Research logoLink to Journal of Radiation Research
. 2013 Jan 7;54(4):597–610. doi: 10.1093/jrr/rrs129

Comparative analysis of radon, thoron and thoron progeny concentration measurements

Miroslaw Janik 1, Shinji Tokonami 2,*, Chutima Kranrod 3, Atsuyuki Sorimachi 2, Tetsuo Ishikawa 1, Masahiro Hosoda 4, James Mclaughlin 5, Byung-Uck Chang 6, Yong Jae Kim 6
PMCID: PMC3709658  PMID: 23297318

Abstract

This study examined correlations between radon, thoron and thoron progeny concentrations based on surveys conducted in several different countries. For this purpose, passive detectors developed or modified by the National Institute of Radiological Sciences (NIRS) were used. Radon and thoron concentrations were measured using passive discriminative radon-thoron detectors. Thoron progeny measurements were conducted using the NIRS-modified detector, originally developed by Zhuo and Iida. Weak correlations were found between radon and thoron as well as between thoron and thoron progeny. The statistical evaluation showed that attention should be paid to the thoron equilibrium factor for calculation of thoron progeny concentrations based on thoron measurements. In addition, this evaluation indicated that radon, thoron and thoron progeny were independent parameters, so it would be difficult to estimate the concentration of one from those of the others.

Keywords: radon, thoron, thoron progeny, thoron equilibrium factor

INTRODUCTION

Inhalation of radon (Rn) and its short-lived decay products, as well as of the thoron (Tn) series, accounts for about half of the effective dose from natural radiation sources. In the past, Rn studies were more common than those of Tn, but recent studies in Serbian [1–3], Chinese [4, 5], Indian [6] Canadian [7–9], Japanese [10], Korean [11, 12], Hungarian [13], Slovenian [14], Irish [15, 16] and USA [17] dwellings have show that Tn can be a significant contributor to the radiation exposure. The contributions of Rn and Tn and its progeny (TnP) to radiation exposure are quite different. The decay half life of Rn is 3.82 d, which is long enough for its transfer from a source (mainly soil) through cracks and gaps into indoor spaces. Several parameters, such as dwelling type, ventilation rate and construction materials, influence the Rn concentration in dwellings. Due to its long half life, Rn is usually well-mixed in room air. On the other hand, due to the relatively short half life of Tn (55.6 s), its presence is highly inhomogeneous in room air and is strongly dependent on distance from the source [18–21]. Also, Tn can interfere with the detection of Rn in detectors that lack diffusion barrier discrimination and thus allow rapid gas entry into the sensitive volume [22]. For a rough estimation of lung dose using the equation given by UNSCEAR [23], several important components are required, such as the equilibrium equivalent concentration (EEC), the equilibrium factor (F) and gas concentration (C).

Since direct measurements of the concentration of all short-lived Rn and Tn decay products are difficult and limited, the EEC is estimated from considerations of equilibrium (or disequilibrium) between Rn and Tn and their decay products, and is expressed by the equation:

graphic file with name rrs129eq1.jpg

where x is Rn or Tn. Several groups have reported the Rn equilibrium factor to be within the range 0.3–0.6 [24–27], which is close to the typical value of 0.4 adopted by UNSCEAR.

On the other hand, simultaneous long-term measurements of Tn and its progeny have rarely been conducted, and there is a shortage of data on EEC for Tn [4, 26–32]. As a result, the progeny concentration is usually estimated by multiplying the Tn concentration by a typical equilibrium factor for Tn (FTn = 0.02), which has also been provided by the UNSCEAR Report.

In the present study, data from six Rn and Tn surveys in China (S1 and S2), Ireland (S3) and Korea (S4, S5, and S6) were analyzed. The Chinese surveys were conducted in Gansu Province by the National Institute of Radiological Science (NIRS, Japan) in cooperation with National Institute for Radiological Protection (NIRP, China) and Kagoshima University (Japan) [4]. A previous study carried out by the Chinese Laboratory of Industrial Hygiene and the US National Cancer Institute showed an increased lung cancer risk due to high residential Rn levels. Other Chinese surveys provided by Tokonami et al. [33] and Sun et al. [34] found areas with high levels of Tn. Simultaneous measurements of indoor Rn, Tn and airborne TnP were made in Ireland for the evaluation of long-term exposures in dwellings [16]. The Korean studies were designed as nationwide surveys for Rn, Tn and TnP to provide annual averages of them and to estimate the effective dose for the general public [35].

All of these surveys were carried out using both the passive Rn and Tn detectors and the TnP monitor developed by the NIRS, and data were examined to investigate any relationship between Rn and Tn; between Rn and TnP; and between Tn and TnP. In addition, the problem of evaluation of the Tn equilibrium factor was considered.

MATERIALS AND METHODS

Rn and Tn measurements

For simultaneous measurements of Rn and Tn, passive integrated Rn–Tn detectors (commercially sold as RADOPOT or RADUET) were used [36–37]. These detectors were developed and evaluated at NIRS. Quality assurance of the detectors was checked by participation in international intercomparisons of passive Rn and Tn detectors; the difference between the reference value and the registered value did not exceed 20% [38–40].

The working principle for both detectors is the same. A detector consists of two diffusion chambers with different air exchange rates. The chamber with the low air exchange rate (LER) is for Rn registration while the high air exchange rate (HER) chamber is for Tn and Rn registration. The Rn gas in the air diffuses to the LER chamber through a very narrow invisible gap, while Tn diffuses through evenly distributed holes made on the HER chamber wall. The differences between RADOPOT and RADUET detectors are mainly: the inner volume of each chamber; the number and diameter of holes; the covering material for the holes of the HER chamber wall; and the materials from which the detectors are made. The lower limit of detection (LLD) is calculated on the basis of an ISO Guideline [41]. The LLD depends on the concentration of both gases and on the exposure period. For example, when a Rn concentration of 15 Bq m−3 and a Tn concentration of 15 Bq m−3 are given with a measurement period of 90 days, the detection limits are estimated to be 5 Bq m−3 and 7 Bq m−3, respectively.

TnP measurement

For the measurement of TnP the detector developed by Zhuo and Iida was used [42]. In the first version of the detector (Fig. 1a) only one CR-39 chip was used. The modified version by NIRS (Fig. 1b), used for these surveys, consists of two or four chips because a larger number of CR-39 chips can achieve better statistics and lower measurement uncertainty. CR-39 chips are covered with an aluminium-vaporized Mylar film of 71-mm air-equivalent thickness. The thickness of the Mylar film allows the detection of only the 8.78 MeV alpha particles emitted from 212Po. The LLD of the modified detector was estimated using the Currie formula [43] to be 0.005 Bq m−3 for an exposure period of 90 days. Based on a laboratory calibration of this TnP monitor, the airborne concentration of the Tn decay product 212Pb in the air was determined and expressed as Bq m−3 equilibrium equivalent thoron concentration (EETC).

Fig 1.

Fig 1.

Open in a new tab

Thoron progeny monitors: (a) developed by Zhuo and Iida [41] and (b) modified by NIRS and used in surveys.

Background and calibration

The background measurements were done using non-exposed detectors, hence registered counts were subtracted from total counts obtained using an exposed detector. In addition, for each series of CR-39 chips a new calibration factor was determined using NIRS Rn and Tn chambers.

The fluctuation of background and efficiency were taken into consideration for estimating Rn, Tn and Tn progeny concentrations [44], and background and efficiency were estimated for each series (product lot) of CR39. The background and efficiency were assumed to be similar within each series (1000 chips of CR39). Even if the background and/or efficiency change from lot to lot, this won't significantly affect the accuracy of estimation. The fluctuation of calibration factors (from 2004) ranges from 0.0088–0.0245 and 0.0049–0.0198 (± 3–5%) tracks mm−2 kBq−1 h−1 m3 for Rn and Tn, respectively, whereas background fluctuation (from 2004) ranges from 0.08–0.60 tracks mm−2.

Software

Statistical evaluations of the data were performed by R-language packages [45–46]. Normality of distributions was checked by Jarque–Bera (J–B) and d'Agostino (d'A) tests. The statistical hypothesis for the Tn equilibrium factor FTn = 0.02 (null hypothesis—recommended value by UNSCEAR) was checked by the Student's t-test if results were normally distributed (based on results given by J–B and d'A tests), whereas the Wilcoxon signed-rank test was used if the results could not be assumed to have a normal distribution. In addition, for quantification of the statistical dependence between any two variables (Rn–Tn, Rn–TnP and Tn–TnP) the linear regression model was applied. The number of classes for histograms was made by the Freedman-Diaconis method based on the inter-quartile range [47]. Moreover, data below LLD were excluded from consideration.

Outliers/hot-spots

Datasets of spatial environmental variables are often found to be approximately log-normally distributed. For geochemical variables the phenomenon is well known, and a lot of literature exists regarding this [48]. In addition, the observation of approximate log-normality is abundant for Rn concentration data, thus the term ‘log-normal mysticism’ has been coined [49], requiring further studies. Considering this situation, we believe that most of the outliers/hot-spots identified by our analysis resulted from ‘anomalous’ sampling procedures: for indoor Rn/Tn measurements, apart from measurement errors and ‘genuinely’ statistical outliers, particular features of the construction of the building, the distance from the wall at which Rn/Tn was measured, behaviour of occupants, environmental conditions, etc. may have led to overestimation of the Rn/Tn concentration. However, the analysis of the nature of anomalies is not within the scope of this study.

In order to identifying the outliers/hot-spots in a survey, the following steps were executed: (i) perform the J–B and d'A tests; (ii) if the P value < 0.05 the highest data was removed, then do the J–B and d'A tests again; and (iii) steps (i) and (ii) were repeated as long as results of the J–B and d'A tests were below the P values.

RESULTS AND DISCUSSION

Results of indoor Rn, Tn and TnP concentration measurements from the six surveys are summarized in Table 1 and shown as box plots in Figs 2a–4a.

Fig 3.

Fig 3.

Open in a new tab

Box plot of thoron concentration results: (a) original data and (b) outlier evaluated data.

Table 1.

Results of measurements: original and outlier evaluated data

Original data
 Outlier evaluated data
S1 S2 S3 S4 S5 S6 S2m S3m S4m S5m S6m
N 104 104 347 387 410 396 89 331 356 402 387
Rn (Bq m−3) Mean 97 81 75 65 68 50 78 75 49 56 42
Stdev 58 40 78 67 157 90 36 80 34 56 34
Median 76 75 50 39 41 33 70 49 37 40 32
GMa,b 82 72 51 – – – 70 52 40 41 34
Min. 16 22 4 4 5 5 22 4 4 5 5
Max. 303 248 767 468 2841 1543 200 767 148 476 283
Skewc 1.18 1.50 3.66 2.83 14.03 12.31 1.10 3.63 1.19 3.41 2.91
Kurtd 0.93 2.99 22.22 10.05 235.03 193.45 1.37 21.59 0.50 16.16 12.33
Tn (Bq m−3) Mean 298 432 25 60 73 44 375 26 52 60 38
Stdev 265 437 25 121 136 74 346 24 70 70 49
Median 206 258 18 34 38 29 246 19 32 38 27
GM 214 – 18 – – – 278 19 32 40 25
Min. 19 47 3 3 3 3 47 3 3 3 3
Max. 1200 1922 189 1996 1829 797 1643 189 697 563 515
Skew 1.83 2 3.11 11.29 7.55 6.12 2.11 3.10 4.80 3.47 5.40
Kurt 2.73 3.01 13.36 167.33 78.95 45.92 3.94 13.31 31.94 15.35 38.55
TnP (Bq m−3) EETC Mean 3.12 2.13 0.46 0.96 1.12 0.78 2.10 0.42 0.96 1.11 0.73
Stdev 1.31 1.03 0.43 0.93 1.19 1.20 0.80 0.29 0.93 1.06 0.91
Median 2.98 2.04 0.35 0.67 0.79 0.45 2.06 0.35 0.69 0.79 0.43
GM 2.85 1.86 – – 0.82 0.44 1.94 0.34 0.36 0.81 0.44
Min. 0.98 0.24 0.01 0.03 0.06 0.02 0.60 0.03 0.03 0.06 0.02
Max. 7.23 5.17 3.3 8.84 7.58 16.45 3.96 1.80 8.84 7.58 6.46
Skew 0.75 0.55 3.35 3.16 3.1 6.86 0.15 1.77 3.22 3.02 3.09
Kurt 0.40 0.02 15.37 16.40 12.57 74.52 -0.81 4.26 17.12 12.32 11.66
FTn Mean 0.019 0.008 0.035 0.038 0.035 0.034 0.009 0.032 0.040 0.035 0.035
Stdev 0.018 0.006 0.05 0.052 0.043 0.055 0.006 0.047 0.053 0.044 0.0.5
Median 0.014 0.008 0.018 0.021 0.021 0.017 0.008 0.018 0.022 0.021 0.018
GM 0.013 0.006 0.018 – – 0.017 0.007 0.018 0.021 0.020 0.017
Min. 0.002 0.001 0.001 0.001 0.001 0.001 0.001 0.001 <0.001 <0.001 0.001
Max. 0.156 0.037 0.5 0.43 0.332 0.573 0.037 0.500 0.430 0.332 0.573
Skew 2.54 1.45 4.05 3.39 3.33 5.09 1.72 4.77 3.29 3.31 5.05
Kurt 8.36 3.79 23.26 15.53 14.62 37.38 5.45 34.42 14.55 14.44 36.77
Tn/Rn ratio Mean 4.33 6.46 0.49 1.38 1.44 1.36 6.17 0.49 1.43 1.38 1.16
Stdev 4.91 6.94 0.42 2.31 1.56 4.12 6.86 0.41 2.26 1.40 1.35
Median 2.34 3.79 0.37 0.77 0.97 0.77 3.56 0.38 0.84 0.97 0.78
GM 2.61 4.2 0.36 – 0.98 – 3.97 0.36 0.82 0.96 0.74
Min. 0.23 0.77 0.04 0.03 0.05 0.03 0.77 0.04 0.05 0.05 0.03
Max. 23.56 38.15 3.05 22.75 13.84 78.25 38.15 3.05 22.75 12.43 14.35
Skew 2.14 2.07 2.4 5.33 3.68 16.56 2.24 2.23 5.40 3.34 4.04
Kurt 4.71 4.36 8.40 35.64 20.42 303.70 5.36 7.38 37.51 18.03 24.94
Open in a new tab

aGeometric mean; bwhen neither normal nor lognormal distributions are appropriate only the mean and median values are given; cskewness; dkurtosis.

Fig 2.

Fig 2.

Open in a new tab

Box plot of radon concentration results: (a) original data (b) outlier evaluated data.

Fig 4.

Fig 4.

Open in a new tab

Box plot of thoron progeny concentration results: (a) original data and (b) outlier evaluated data.

All of the surveys were focused on characterizing Tn and TnP levels, and were analyzed to gauge the interference that Tn may cause in typical Rn measurements performed for home risk assessment. For calculation purposes, only raw data with triple results (Rn, Tn and TnP) obtained simultaneously were considered. Data from the different surveys were not merged for analysis due to different detector mounting places. In surveys S1 and S2, RADOPOT detectors were placed at a distance from the wall/ceiling (range, 5–30 cm), whereas for surveys S3–S6 RADUET detectors were mounted on the wall or another suitable surface in the living room. As mentioned earlier, the Tn concentration strongly depends on the distance from the wall. Therefore, if the detector is mounted directly on a wall, it gives an accurate determination of TnP in the room and an indication of the maximum Tn concentration. Furthermore, if the detector distance from the wall is more than several centimetres the Tn concentration measured can be assumed to be representative of the room.

The range of averages of indoor Rn, Tn and TnP are 50–97 Bq m−3, 25–432 Bq m−3 and 0.46–3.12 Bq m−3 (EETC), respectively.

In the Chinese surveys (S1 and S2), detectors were suspended from the ceiling in the centre of each cave dwelling (range, 5–30 cm), while they were hung on walls for other surveys. The cave dwelling is a particular form of earth shelter dwelling common in the Loess Plateau in China's north. It consists of one room with a single entrance and two windows at the front side. The length is 8–10 m with a width of 3–3.5 m and a height of 3–3.5 m. The cave dwelling is equipped with a traditional bed formed from a loess cube, which is called Kang in Chinese. Regardless of the placement of detectors with some distance from the ceiling and walls (possible source of Tn) in Chinese surveys, the averages of Tn concentration are generally higher. This could be because the cave dwellings were mainly made from the loess without wall coverings and walls (ceilings) are strong sources of Tn. In other surveys, the investigated houses were made from typical building materials with painted walls, and consequently the soil under the houses could be considered the main source of Tn.

Because the average and median values were different in all cases, the skewness for measuring the asymmetry of the distributions was calculated and the results are shown in Table 1. All original (S1–S6) cases show positive skewness (the right tail is longer). This means that values are concentrated mainly on the left side and high values of the right side disturb (increase) the average value. These high values can be recognized as potential outliers/hot-spots and analyzed later.

The J–B and d'A normality tests (Table 2) show that some of the distributions of original data are log-normal at the 0.05 significance level.

Table 2.

Jarque–Bera (J–B) and d'Agostino (d'A) normality tests results

Original data
 Outlier evaluated data
Survey Parameter S1 S2 S3 S4 S5 S6 S2m S3m S4m S5m S6m
J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A J–B d'A
Ln(Rn) Oa / O O / O O / O - / - - / - - / - O / O O / O O / O O / O O / O
Ln(Tn) O / O - / - O / O - / - - / - - / - O / O O / O O / O O / O O / O
Ln(TnP) O / O - / - - / - O / O O / O O / O O / O O / O O / O O / O O / O
Ln(F) O / O - / - O / O O / O - / - O / O O / O O / O - / - O / O O / O
Ln(Tn/Rn) O / -b O / - O / O O / - O / - - / - O / O O / O O / - O / O O / O
Open in a new tab

aO : passed; b- : failed

The ranges of skewness for S3–S6 Tn measurements are greater than for S1 and S2, however the average values of the concentrations are lower. The difference can be explained by the mounting place of the detectors (for S3–S6 close to the wall, and for S1 and S2 at a distance from the wall), therefore S3–S6 detectors could register close to the maximum Tn concentrations in the rooms. This indicates that special attention must be paid to the location of the detector if a single device is to be used for further evaluation and dose estimation.

In most cases if the survey is conducted within an area up to a few tens of kilometre radius, indoor Rn concentration has a high probability of being log-normally distributed [47]. Based on this assumption a simple method using normality tests has been implemented, which filters outliers from a data population that is known to be (approximately) log-normally distributed.

All the surveys considered in this paper were carried out locally within the maximum 30 km radius, but only a few distributions of Rn, Tn and TnP passed normality tests (Table 2). One of the reasons for failing the tests is because these surveys were investigated using a single set of detectors (Rn–Tn–TnP) therefore some of results may be inconsistent with other observations in the data population and clearly visible in the Q-Q plots and histograms. Possible outliers are enclosed by ovals in Figs 5a–7a. Other potential sources of error, such as environmental parameters (e.g. pressure and temperature gradients) and ventilation rate, were not analyzed in this paper. Therefore, if the original data did not pass the normality tests, the outlier evaluation to find outliers (or anomalies or hot spots) was executed, i.e. removal of the maximum values was stopped if the normality tests (J–B and d'A) were passed. The modified results (excluding S1) are denoted as S2m for the S2 survey, … S6m for the S6 survey, and are listed in Table 1 and presented as box plots in Figs 2b–4b. Because Rn, Tn and TnP for the S1 survey passed the normality tests, no outlier evaluation for this series was performed.

Fig 6.

Fig 6.

Open in a new tab

Histograms and Q-Q plot of thoron concentration for S5 survey: (a) original data with indication of possible outliers by ovals and (b) outlier evaluated data (histogram with log-normal distribution).

Fig 5.

Fig 5.

Open in a new tab

Histograms and Q-Q plot of radon concentration for S5 survey: (a) original data with indication of possible outliers by ovals and (b) outlier evaluated data (histogram with log-normal distribution).

Fig 7.

Fig 7.

Open in a new tab

Histograms and Q-Q Plots of thoron progeny concentration for S6 survey: (a) original data with indication of possible outliers by ovals and (b) outlier evaluated data (histogram with log-normal distribution).

It should be noticed that after outlier evaluation all data for Rn, Tn and TnP concentrations showed log-normal distributions (Table 2 and Figs 5b–7b).

Because one of the most important values for dose calculation is the TnP concentration, the linear dependency between Tn and its progeny was calculated. The results of the calculation are presented in Fig. 8 (as an example) and in Table 3. It was found that the coefficient of determination (R2) between all surveys is very weak and the R2 parameter is low, being below 0.05.

Fig 8.

Fig 8.

Open in a new tab

Correlation relationship between thoron and thoron progeny concentrations for S5 survey: (a) original data, (b) outlier evaluated data.

Table 3.

Coefficients of determination R2 for correlation between Rn, Tn and TnP results

Original data
 Outlier evaluated data
Survey Parameter S1 S2 S3 S4 S5 S6 S2m S3m S4m S5m S6m
Rn–Tn <0.01 0.01 0.43 0.03 0.48 0.12 0.01 0.44 0.01 0.19 0.16
Rn–TnP <0.01 0.03 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01
Tn–TnP 0.04 0.01 <0.01 0.03 0.02 0.02 0.05 <0.01 <0.01 0.01 0.04
Open in a new tab

In addition, the linear regression model was adopted for other measured parameters and the results are summarized in Table 3. As expected, the relationship between Rn and Tn is weak, with the coefficient of determination < 0.2 (with S3 and S5 being exceptions). The regression analysis also shows a poor correlation between Rn and TnP (R2 < 0.03). These results suggest that the concentrations appear to be independent of each other.

Another important parameter for dose calculation is equilibrium factor (FTn). Results of the FTn calculation, defined as the ratio of TnP/Tn, are listed in Table 1. The range of mean values of this parameter for all data vary from <0.01 to 0.04, with the median of <0.01 to 0.02 and a range from <0.01 to 0.57. It should be noted that UNSCEAR 2000 assumed average equilibrium factors (FTn) for Tn obtained from the 212Pb/220Rn ratio to be 0.02 indoors and 0.003 outdoors. As mentioned earlier the appropriate tests for normal and non-normal distributions were performed. Results of these calculations are shown in Table 4. If the distribution of FTn could be assumed to be log-normal (Table 2) the t-test was performed, otherwise the Wilcoxon test was performed; if the P-value was < 0.05, the null hypothesis (FTn = 0.02) should be rejected. Only one set (S1) of original data and one set of modified data (S2m) show a positive result, P > 0.05.

Table 4.

Results of Wilcoxon-test and t-test for FTn = 0.02–recommended value by UNSCEAR 2000 [23]

Original data
 Outlier evaluated data
Survey Parameter S1 S2 S3 S4 S5 S6 All S1m S2m S3m S4m S5m S6m All m
Wilcoxon –b – – – –
t-test Oa – – – – O O – –
Open in a new tab

aO : passed; b–: failed

Generally, the influence of Tn on Rn measurement is negligible due to its small value relative to that of Rn and to its short half-life time. But, in some cases the Tn concentration is much higher than that of Rn therefore more accurate evaluation of Rn is necessary. Earlier investigations of passive type detectors have shown the influence of Tn on Rn evaluation. For instance, Tokonami et al. [22] and Sugino et al. [10] showed that if the ratio of Tn/Rn exceeded 1.0, Rn concentration might by overestimated by about 4% by using a single Rn monitor, however a ratio of about 100 could modify results by a factor of 10. On the other hand, a ratio below 0.8 could not modify the Rn results. Moreover, some kinds of detectors tested by Tokonami et al. and Sugino et al. might provide a higher value of Rn than its actual concentration if the influence of Tn is not considered, especially close to the Tn source (wall, ceiling, etc.). As mentioned earlier, for all six surveys studied discriminating Rn and Tn detectors were used. The calculated average Tn/Rn ratios ranged from 0.48–6.46, and the median from 0.37–3.79, with the absolute ranges from 0.03–78.25. In surveys S1, S2 and S5 the average concentrations of Tn were higher than Rn. Analysis of the data shows cumulative relative frequency distribution of ratios <0.8 from 1 to 83% (original data: S1–12%, S2–1%, S3–83%, S4–51%, S5–41%, S6–52%, and modified data: S2m–1%, S3m–82%, S4–48%, S5–41%, S6–52%) with 49% for the original data and 50% for all complied outlier evaluated data (Fig. 9b). The smallest values are obtained in the S1 and S2 surveys, dedicated to Tn investigation.

Fig 9.

Fig 9.

Open in a new tab

Histograms for FTn and Tn/Rn ratio for outlier evaluated data compiled from all six surveys.

The above results, especially for the Tn equilibrium factor, suggest that in most cases TnP concentrations cannot be estimated from the observed concentration of Tn by assuming the Tn equilibrium factor provided by UNSCEAR for individual surveys. The effective dose can be estimated using the following equation:

graphic file with name rrs129eq1.jpg

where HTn is the annual effective dose for Tn decay products (mSv y−1), EECTn is the equilibrium equivalent Tn concentration (Bq m−3), t is the indoor exposure time (usually = 7000 h), DCFTn is the dose conversion factor for Tn (= 40 nSv h−1 Bq−1 m3), and CTn and FTn are as defined earlier. The above equation is very sensitive to FTn factor, e.g. for CTn = 100 Bq m−3 and FTn = 0.008 (minimum value from Table 1) the effective dose is HTn = 0.24 mSv y−1, but for FTn = 0.038 (maximum value) the calculated effective dose is about five times higher (HTn = 1.06 mSv y−1). Therefore, if discriminative measurements are made without incorporating EECTn data, the final effective dose might be over- or under- estimated.

Moreover, for the all merged data the FTn mean value is 0.032 with the range from 0.001–0.573 (with standard deviation of 0.047 and median value of 0.018). The low P-value (<0.05) of the FTn t-test for merged data (the histogram of all six compiled surveys is presented in Fig. 9a) suggests that the UNSCEAR value should be revised if it is to be used for general applications and as a point of reference. On the other hand, the high variation of the equilibrium factor (by a factor of 103) as well as the high value of the standard deviation (more than 100% amount of mean value) may result in highly uncertain risk estimates of the effective dose whenever the FTn value is assumed to be that given by UNSCEAR.

CONCLUSION

A total of 1748 cases were analyzed from six surveys in different parts of the world. In no case was any linear relationship between Tn and its progeny concentrations found. Other analyzed parameters (Rn–Tn and Rn–TnP) also showed weak correlations but some exceptions (stronger correlations) were found.

From the viewpoint of dose assessment TnP measurement is important. However, it is difficult to estimate its concentration from Tn measurement and the typical Tn equilibrium factor as recommended by UNSCEAR. This is because the distribution of Tn concentration within a room varies quite strongly spatially and depends on the distance from its source. Moreover, high values (both mean and range) of the Tn/Rn ratio suggested that the results of Rn measurements without Rn–Tn discrimination might be overestimated if the detector is sensitive to Tn and the measurement is made by devices with no Rn–Tn discrimination capability. It was concluded from the various surveys that, for good estimation of the Rn and Tn dose, measurements of Rn, Tn and their progeny concentrations should be carried out simultaneously.

FUNDING

This work was partially supported by the grant-in-aid ‘Construction of Natural Radiation Exposure Study Network’ from the Special Coordination Funds for Promoting Science and Technology of the Ministry of Education, Culture, Sports, Science and Technology, Japan.

ACKNOWLEDGEMENTS

The authors thank M. Murray from the Radiological Protection Institute of Ireland (RPII) for providing the Irish data for evaluation.

REFERENCES

  • 1.Žunić ZS, Kobal I, Vaupotič J, et al. High natural radiation exposure in radon spa areas: a detailed field investigation in Niška Banja (Balkan region) J Environ Radioactiv. 2006;89:249–60. doi: 10.1016/j.jenvrad.2006.05.010. [DOI] [PubMed] [Google Scholar]
  • 2.Žunić ZS, Kozak K, Ciotoli G, et al. A campaign of discrete radon concentration measurements in soil of Niška Banja town, Serbia. Radiat Meas. 2007;42:1696–702. [Google Scholar]
  • 3.Žunić ZS, Janik M, Tokonami S, et al. Field experience with soil gas mapping using Japanese passive radon/thoron discriminative detectors for comparing high and low radiation areas in Serbia (Balkan Region). J. Radiat Res. 2009;50:355–61. doi: 10.1269/jrr.08098. [DOI] [PubMed] [Google Scholar]
  • 4.Yamada Y, Sun Q, Tokonami S, et al. Radon-thoron discriminative measurements in Gansu province, China, and their implication for dose estimates. J Toxicol Env Heal A. 2006;69:723–34. doi: 10.1080/15287390500261265. [DOI] [PubMed] [Google Scholar]
  • 5.Tokonami S, Sun Q, Akiba S, et al. Radon and thoron exposures for cave residents in Shanxi and Shaanxi Provinces. Radiat Res. 2004;162:390–6. doi: 10.1667/rr3237. [DOI] [PubMed] [Google Scholar]
  • 6.Ramola RC, Prasad G, Gusain GS, et al. Preliminary indoor thoron measurements in high radiation background area of south-eastern coastal Orissa, India. Radiat Prot Dosim. 2010;141:379–82. doi: 10.1093/rpd/ncq238. [DOI] [PubMed] [Google Scholar]
  • 7.Chen J, Tokonami S, Sorimachi A, et al. Preliminary results of simultaneous radon and thoron tests in Ottawa. Radiat Prot Dosim. 2008;130:253–6. doi: 10.1093/rpd/ncm503. [DOI] [PubMed] [Google Scholar]
  • 8.Chen J, Schroth E, MacKinlay E, et al. Simultaneous Rn-222 and Rn-220 measurements in Winnipeg, Canada. Radiat Prot Dosim. 2009;134:75–8. doi: 10.1093/rpd/ncp065. [DOI] [PubMed] [Google Scholar]
  • 9.Chen J, Walker B, Sorimachi A, et al. An investigation on radon and thoron response of alpha-track detectors used in the Winnipeg case–control study. Radiat Prot Dosim. 2010;138:83–6. doi: 10.1093/rpd/ncp177. [DOI] [PubMed] [Google Scholar]
  • 10.Sugino M, Tokonami S, Zhuo W. Radon and thoron concentrations in offices and dwellings of the Gunma prefecture, Japan. J Radioanal Nucl Chem. 2005;266:205–9. [Google Scholar]
  • 11.Kim CK, Kim YJ, Lee HY, et al. Rn-220 and its progeny in dwellings of Korea. Rad Meas. 2007;42:1409–14. [Google Scholar]
  • 12.Chung W, Tokonami S. Preliminary survey on radon and thoron concentrations in Korea. Radiat Prot Dosim. 1998;80:423–6. [Google Scholar]
  • 13.Kávási N, Németh C, Kovács T, et al. Radon and thoron parallel measurements in Hungary. Radiat Prot Dosim. 2007;123:250–3. doi: 10.1093/rpd/ncl102. [DOI] [PubMed] [Google Scholar]
  • 14.Vaupotič J, Kávási N. Preliminary study of thoron and radon levels in various indoor environments in Slovenia. Radiat Prot Dosim. 2010;141:383–5. doi: 10.1093/rpd/ncq251. [DOI] [PubMed] [Google Scholar]
  • 15.McLaughlin J. An overview of thoron and its progeny in the indoor environment. Radiat Prot Dosim. 2010;141:316–21. doi: 10.1093/rpd/ncq234. [DOI] [PubMed] [Google Scholar]
  • 16.McLaughlin J, Murray M, Currivan L, et al. Long-term measurements of thoron, its airborne progeny and radon in 205 dwellings in Ireland. Radiat Prot Dosim. 2011;145:189–93. doi: 10.1093/rpd/ncr067. [DOI] [PubMed] [Google Scholar]
  • 17.Harley N, Chittaporn P, Medora R, et al. Measurement of the indoor and outdoor 220Rn (thoron) equilibrium factor: application to lung dose. Radiat Prot Dosim. 2010;141:357–62. doi: 10.1093/rpd/ncq228. [DOI] [PubMed] [Google Scholar]
  • 18.Doi M, Kobayashi S. Vertical distribution of outdoor radon and thoron in Japan using a new discriminative dosimeter. Health Phys. 1994;67:385–92. doi: 10.1097/00004032-199410000-00009. [DOI] [PubMed] [Google Scholar]
  • 19.Gargioni E, Model R. A simple method for measuring thoron spatial distributions. Radiat Prot Dosim. 2005;113:321–5. doi: 10.1093/rpd/nch467. [DOI] [PubMed] [Google Scholar]
  • 20.Tokonami S. Why is Rn-220 (thoron) measurement important? Radiat Prot Dosim. 2010;141:335–9. doi: 10.1093/rpd/ncq246. [DOI] [PubMed] [Google Scholar]
  • 21.Doi M, Fujimoto K, Kobayashi S, et al. Spatial distribution of thoron and radon concentrations in the indoor air of a traditional Japanese wooden house. Health Phys. 1994;66:43–9. doi: 10.1097/00004032-199401000-00006. [DOI] [PubMed] [Google Scholar]
  • 22.Tokonami S, Yang M, Sanada T. Contribution from thoron on the response of passive radon detectors. Health Phys. 2001;80:612–5. doi: 10.1097/00004032-200106000-00014. [DOI] [PubMed] [Google Scholar]
  • 23.UNSCEAR. New York: United Nation Scientific Committee on the Effects of Atomic Radiation; 2000. Sources and effects of ionizing radiation. [Google Scholar]
  • 24.Clouvas A, Xanthos S, Antonopoulos-Domis M. Simultaneous measurements of indoor radon, radon–thoron progeny and high-resolution gamma spectrometry in Greek dwellings. Radiat Prot Dosim. 2006;118:482–90. doi: 10.1093/rpd/nci366. [DOI] [PubMed] [Google Scholar]
  • 25.Chen C-J, Liu C-C, Lin Y-M. Measurement of equilibrium factor and unattached fraction of radon progeny in Kaohsiung, Taiwan. Appl Radiat Isot. 1998;49:1613–8. doi: 10.1016/s0969-8043(98)00065-7. [DOI] [PubMed] [Google Scholar]
  • 26.Yu KN, Leung SYY, Nikezic D, et al. Equilibrium factor determination using SSNTDs. Radiat Meas. 2008;43:357–63. [Google Scholar]
  • 27.Tokonami S, Furukawa M, Shicchi Y, et al. Characteristics of radon and its progeny concentrations in air-conditioned office buildings in Tokyo. Radiat Prot Dosim. 2003;106:71–6. doi: 10.1093/oxfordjournals.rpd.a006338. [DOI] [PubMed] [Google Scholar]
  • 28.Ramola RC, Negi MS, Choubey VM. Measurement of equilibrium factor “F” between radon and its progeny and thoron and its progeny in the indoor atmosphere using nuclear track detectors. Indoor Built Environ. 2003;12:351–5. [Google Scholar]
  • 29.Mishra R, Mayya YS. Study of a deposition-based direct thoron progeny sensor (DTPS) technique for estimating equilibrium equivalent thoron concentration (EETC) in indoor environment. Radiat Meas. 2008;43:1408–16. [Google Scholar]
  • 30.Chen J, Moir D, Sorimachi A, et al. Characteristics of thoron and thoron progeny in Canadian homes. Radiat Environ Biophys. 2011;50:85–9. doi: 10.1007/s00411-010-0338-5. [DOI] [PubMed] [Google Scholar]
  • 31.Chen J, Moir D, Pronk T, et al. An update on thoron exposure in Canada with simultaneous 222Rn and 220Rn measurements in Fredericton and Halifax. Radiat Prot Dosim. 2011;147:541–7. doi: 10.1093/rpd/ncq567. [DOI] [PubMed] [Google Scholar]
  • 32.Chen J, Moir D, Sorimachi A, et al. Determination of thoron equilibrium factor from simultaneous long-term thoron and its progeny measurements. Radiat Prot Dosim. 2012;149:155–8. doi: 10.1093/rpd/ncr228. [DOI] [PubMed] [Google Scholar]
  • 33.Tokonami S, Sun Q, Akiba S, et al. Natural radiation exposures for cave residents in China. In: Simopoulos ES, editor. Radioactivity in the Environment. Elsevier: Amsterdam; 2005. pp. 560–6. [Google Scholar]
  • 34.Sun Q, Tokonami S, Hou C, et al. Epidemiological potentials of radon- and thoron-prone area in China. Jpn J Health Phys. 2004;39:257–62. [Google Scholar]
  • 35.Kim YJ, Lee HY, Kim CS, et al. Indoor radon, thoron, and thoron daughter concentrations in Korea. Int Congr Ser. 2005;1276:46–9. [Google Scholar]
  • 36.Zhuo W, Tokonami S, Yonehara H, et al. A simple passive monitor for integrating measurements of indoor thoron concentrations. Rev Sci Instrum. 2002;73:2877–81. [Google Scholar]
  • 37.Tokonami S, Takahashi H, Kobayashi Y, et al. Up-to-date radon-thoron discriminative detector for a large scale survey. Rev Sci Instrum. 2005;76:113505–9. [Google Scholar]
  • 38.Janik M, Tokonami S, Kovács T, et al. International intercomparisons of integrating radon detectors in the NIRS radon chamber. Appl Radiat Isot. 2009;67:1691–6. doi: 10.1016/j.apradiso.2009.03.006. [DOI] [PubMed] [Google Scholar]
  • 39.Janik M, Tokonami S, Kranrod C, et al. International intercomparisons of integrating radon/thoron detectors with the NIRS radon/thoron chambers. Radiat Prot Dosim. 2010;141:436–9. doi: 10.1093/rpd/ncq230. [DOI] [PubMed] [Google Scholar]
  • 40.Vargas A, Ortega X. Influence of environmental changes on integrating radon detectors: results of an intercomparison exercise. Radiat Prot Dosim. 2007;123:529–36. doi: 10.1093/rpd/ncl161. [DOI] [PubMed] [Google Scholar]
  • 41.International Organization for Standardization (ISO) Geneva, Switzerland, 2010: International Organization for Standardization; Determination of the characteristic limits (decision threshold, detection limit and limits of the confidence interval) for measurements of ionizing radiation: fundamentals and application. ISO 11929. [Google Scholar]
  • 42.Zhuo W, Iida T. Estimation of thoron progeny concentrations in dwellings with their deposition rate measurements. J Health Phys. 2000;35:365–70. [Google Scholar]
  • 43.Currie LA. Limits for qualitative detection and quantitative determination. Application to radiochemistry. Anal Chem. 1968;40:586–93. [Google Scholar]
  • 44.Tokonami S, Zhuo W, Ryuo H, et al. Instrument performance of a radon measuring system with the alpha-track detection technique. Radiat Prot Dosim. 2003;103:69–72. doi: 10.1093/oxfordjournals.rpd.a006118. [DOI] [PubMed] [Google Scholar]
  • 45.R Development Core Team. R. A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing; 2010. [Google Scholar]
  • 46.Crawley MJ. London: John Wiley & Sons, Ltd; 2010. Statistics: an introduction using R. [Google Scholar]
  • 47.Freedman D, Diaconis P. On the histogram as a density estimator: L2 theory. Probab Theory Rel. 1981;57:453–76. [Google Scholar]
  • 48.Bossew P. Radon: exploring the log-normal mystery. J. Environ Radioact. 2010;101:826–34. doi: 10.1016/j.jenvrad.2010.05.005. [DOI] [PubMed] [Google Scholar]
  • 49.Tóth E, Hámori K, Minda M, et al. Indoor radon in Hungary (lognormal mysticism) In: Barnet I, Neznal M, Pacherová P, editors. Proceedings of Radon Investigations in the Czech Republic XI and the 8th International Workshop on the Geological Aspects of Radon Risk Mapping. Prague: 2006. [Google Scholar]

Articles from Journal of Radiation Research are provided here courtesy of Oxford University Press

RESOURCES