Invited article: in situ comparison of passive radon-thoron discriminative monitors at subsurface workplaces in Hungary.

Invited Article: In situ comparison of passive radon-thoron discriminative monitors at subsurface workplaces in Hungary Norbert Kávási, Tamás Vigh, Csaba Németh, Tetsuo Ishikawa, Yasutaka Omori, Miroslaw Janik, and Hidenori Yonehara Citation: Review of Scientific Instruments 85, 022002 (2014); doi: 10.1063/1.4865161 View online: http://dx.doi.org/10.1063/1.4865161 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/2?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Invited Article: Radon and thoron intercomparison experiments for integrated monitors at NIRS, Japan Rev. Sci. Instrum. 85, 022001 (2014); 10.1063/1.4865159 A transportable spectrometer for in situ and local measurements of iodine monoxide at mixing ratios in the 1014 range Appl. Phys. Lett. 100, 251110 (2012); 10.1063/1.4726190 Principle, calibration, and application of the in situ alkali chloride monitor Rev. Sci. Instrum. 80, 023104 (2009); 10.1063/1.3081015 A broadband absorption spectrometer using light emitting diodes for ultrasensitive, in situ trace gas detection Rev. Sci. Instrum. 79, 123110 (2008); 10.1063/1.3046282 A low cost microsatellite instrument for the in situ measurement of orbital atomic oxygen effects Rev. Sci. Instrum. 68, 3220 (1997); 10.1063/1.1148270

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 89.233.219.231 On: Tue, 08 Apr 2014 16:41:46

REVIEW OF SCIENTIFIC INSTRUMENTS 85, 022002 (2014)

Invited Article: In situ comparison of passive radon-thoron discriminative monitors at subsurface workplaces in Hungary Norbert Kávási,1,2,a) Tamás Vigh,2,3 Csaba Németh,2,4 Tetsuo Ishikawa,1 Yasutaka Omori,1 Miroslaw Janik,1 and Hidenori Yonehara1 1

National Institute of Radiological Sciences, Chiba, Japan Social Organization for Radioecological Cleanliness, Veszprém, Hungary 3 Manganese Mining Process Ltd., Úrkút, Hungary 4 University of Pannonia, Veszprém, Hungary 2

(Received 12 March 2013; accepted 3 November 2013; published online 19 February 2014) During a one-year long measurement period, radon and thoron data obtained by two different passive radon-thoron discriminative monitors were compared at subsurface workplaces in Hungary, such as mines (bauxite and manganese ore) and caves (medical and touristic). These workplaces have special environmental conditions, such as, stable and high relative humidity (100%), relatively stable temperature (12◦ C–21◦ C), low or high wind speed (max. 2.4 m s−1 ) and low or elevated aerosol concentration (130–60 000 particles m−3 ). The measured radon and thoron concentrations fluctuated in a wide range among the different workplaces. The respective annual average radon concentrations and their standard deviations (in brackets) measured by the passive radon-thoron discriminative monitor with cellulose filter (CF) and the passive radon-thoron discriminative monitor with sponge filter (SF) were: 350(321) Bq m−3 and 550(497) Bq m−3 in the bauxite mine; 887(604) Bq m−3 and 1258(788) Bq m−3 in the manganese ore mine; 2510(2341) Bq m−3 and 3403(3075) Bq m−3 in the medical cave (Hospital Cave of Tapolca); and 6239(2057) Bq m−3 and 8512(1955) Bq m−3 in the touristic cave (Lake Cave of Tapolca). The respective average thoron concentrations and their standard deviation (in brackets) measured by CF and SF monitors were: 154(210) Bq m−3 and 161(148) Bq m−3 in the bauxite mine; 187(191) Bq m−3 and 117(147) Bq m−3 in the manganese-ore mine; 360(524) Bq m−3 and 371(789) Bq m−3 in the medical cave (Hospital Cave of Tapolca); and 1420(1184) Bq m−3 and 1462(3655) Bq m−3 in the touristic cave (Lake Cave of Tapolca). Under these circumstances, comparison of the radon data for the SF and CF monitors showed the former were consistently 51% higher in the bauxite mine, 38% higher in the manganese ore mine, and 34% higher in the caves. Consequently, correction is required on previously obtained radon data acquired by CF monitors at subsurface workplaces to gain comparable data for SF monitors. In the case of thoron, the data were unreliable and no significant tendency was seen during the comparison therefore comparison of previously obtained thoron data acquired by either CF or SF is doubtful. There was probable influence by relative humidity on the detection response; however, the effects of the high wind speed and elevated aerosol concentration could not be excluded. The results of this study call attention to the importance of calibration under extreme environmental conditions and the need for using reliable radon-thoron monitors for subsurface workplaces. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4865161] I. INTRODUCTION

During recent decades, health problems caused by radon (222 Rn) and its progeny has drawn attention. Consequently, indoor radon surveys have now been made in 67 countries worldwide, covering 5 × 109 people. Radon surveys to identify radiation exposure in specific populations are also currently in progress.1 The presence of radon and its progeny may cause not only health but even measuring protocol issues, hence it is necessary to consider difficulties in measuring processes owing to their temporal variations (daily, seasonal).2–4 In the 1980s, thoron (another isotope of radon – 220 Rn) risk management was a specific concern for uranium mining communities as the presence of increased thoron concena) [email protected]

0034-6748/2014/85(2)/022002/14/$30.00

trations was revealed in subsurface mining operations that resulted from difficulties in dosimetry and radon progeny measurement.5, 6 Until the beginning of the 2000s, indoor measurements of thoron had not got much attention due to its short half-life (55.6 s), which means no considerable accumulation would be expected in indoor air; however, elevated thoron levels were discovered in India, Japan, and China.7–14 Thoron, a radioactive noble gas, occurs in the natural decay chain headed by thorium (232 Th). Due to its short halflife thoron exhalation is only affected by thorium present in depths of a few centimetres, for both building materials and soil. The soil layer affecting exhalation may be 1.5–2.5 m for radon, which means a much greater absolute activity of the source element (238 U(226 Ra)) than that occurring in a few centimetres thick layer.15, 16 Consequently, the accumulation of thoron to the same extent as radon is rather rare, although it is not impossible.

85, 022002-1

© 2014 AIP Publishing LLC


022002-2

Kávási et al.

Historically, attitudes to health risks resulting from the inhalation of thoron and its progeny has changed over time. Since 1955, in the mining industry the acceptable progeny level had been measured as working level (WL). This unit is 1.3 × 105 MeV of the potential alpha energy released by any combination of short-lived radon progeny (218 Po, 214 Pb, 214 Bi) or thoron progeny (212 Pb, 212 Bi) in 1 l of air. One WL equals 3700 Bq m−3 equilibrium equivalent concentration (EEC) for radon and 277.5 Bq m−3 for thoron. This means there is about 13 times higher health risk to thoron progeny relative to the same concentration level of radon progeny.17 On the other hand in recent years, the UNSCEAR reported a four times higher risk to thoron progeny (EEC) involved in the dose conversion factor (40 nSv Bq−1 h−1 m3 derived from a dosimetric model against 9 nSv Bq−1 h−1 m3 derived from epidemiological studies), hence there are decrement tendency in risk values for thoron and its progeny.18 However, thoron has not come to the foreground lately due to its health effects, but rather due to its effects on the results obtained with passive radon measuring instruments in large scale studies.19, 20 Thoron is not distributed evenly in a closed space; due to its short half-life it is mainly found close to its source. Since building materials usually contain thorium (228 Th(224 Ra)), detectable thoron concentrations may occur near building walls. This can unfavourably affect radon measurements when passive radon measuring devices typically used for radon surveys are placed on or near walls and the diffusion chambers of the radon measurement devices do not prevent thoron from entering the chambers.20, 21 This issue has led to the inspection of thoron penetrability into diffusion chambers and also to the improvement of passive radon-thoron measuring instruments.19 Thoron surveys in Hungary have been done using passive radon-thoron discriminative monitors, which were calibrated at the National Institute of Radiological Sciences (NIRS), Japan.22–24 During previous surveys the first prototype, the passive radon-thoron discriminative monitor with cellulose filter (CF) was used,13, 21, 25 but nowadays its improved version, the passive radon-thoron discriminative monitor with sponge filter (SF) is being widely used in many countries not only in Hungary.26–33 These CF and SF passive measuring instruments were designed to measure the radon and thoron average concentrations of residential areas; however, they have also been used at workplaces and underground areas as well.31, 34 In connection with this, the question arises about how comparable the results of the two types of monitors are considering the extreme circumstances (humidity, aerosol concentration, and wind speed) encountered in underground workplaces. To clarify this issue, a one-year long comparative study under different environmental conditions, such as aerosol concentration and air exchange rate (wind speed), was carried out at four underground workplaces in Hungary. The detectors, which were changed monthly, were deployed at eight points in a bauxite mine, at ten points in a manganese ore mine, at three points in a cave with a lake (touristic facility), and at three points in a hospital cave (medical facility).

Rev. Sci. Instrum. 85, 022002 (2014)

II. MATERIALS AND METHODS A. Measuring devices

1. Environmental parameter monitors

Temperature and humidity measurements with TD T&R72Ui Temp/Humidity Data Logger, wind speed measurement with SIBATA Wind Boy ISA-80 thermal anemometer and aerosol concentration measurement in the particle range of 20–1000 nm with P-TRAK Model 8525 were carried out. 2. CF and SF monitors

Figs. 1 and 2 show drawings of the CF and SF radonthoron discriminative monitors, respectively. Each contains two different diffusion chambers, one with a low exchange rate and the other with a high exchange rate. For detection of radon and thoron concentrations allyl diglycol carbonate (CR-39) detectors (solid state nuclear track detectors) are used, and they are fixed on the centre of the chamber bottom using a small piece of removable adhesive material. During an exposure period, radon in air can penetrate into the low exchange rate and high exchange rate diffusion chambers through the invisible air gaps between their lids and bottoms through diffusion, as well as through the filter paper and sponge used in the high exchange rate diffusion

FIG. 1. Schematic drawing of the CF monitor (unit: mm). Reprinted with permission from Zhuo et al., Rev. Sci. Instrum. 73, 2877 (2002). Copyright 2002 American Institute of Physics.


022002-3

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

FIG. 2. Schematic drawing of the SF monitor (unit: mm). Reprinted with permission from Tokonami et al., Rev. Sci. Instrum. 76, 113505 (2005). Copyright 2005 American Institute of Physics.

chamber. Thoron in air can penetrate into just the high exchange rate diffusion chamber as its entry into the low exchange rate chamber is prevented due to its shorter half life compared to radon.35, 36 After the exposure period, both CR-39 detectors are removed from the chambers and chemically etched; then alpha particle tracks on the surface of the detectors are counted with a microscope-mounted track reading system. The average radon and thoron concentrations during the exposure can be calculated using two alpha particle track densities of the low and high air exchange rate chambers (TDL and TDH in tracks of cm−2 ) with the next two equations:35 T DL = CRn × CfRn1 × t + CT n × CfT n1 × t + B,

(1)

T DH = CRn × CfRn2 × t + CT n × CfT n2 × t + B,

(2)

where CRn and CTn are the average concentrations of radon and thoron during the exposure in Bq m−3 , CfRn1 , and CfTn1 are the radon and thoron calibration factors for the low air exchange rate chamber in tracks of cm−2 kBq−1 m3 h−1 , CfRn2 , and CfTn2 are the radon and thoron calibration factors for the high air exchange rate chamber in tracks of cm−2 kBq−1 m3 h−1 , t is the exposure time in hours, and B is the background alpha particle track density on the

CR-39 detector in tracks of cm−2 . The main specifications of the monitors are summarized in Tables I and II.

3. Calibration of the monitors

The radon and thoron calibration factors for the discriminative monitors are obtained from a calibration process done using calibration chambers at NIRS. To produce the

TABLE I. Main specifications of CF monitor. Detector position Material Filter Dimensions

Radon exposure Thoron exposure


1 chip fastened under the lid and 1 chip on the bottom Polypropylene Air gap, cellulose filter Cylindrical, φ35 mm X 55 mm Cf for low exchange (tracks of cm−2 kBq−1 m3 h−1 ) 1.65 ± 0.15 0.01 ± 0.0009 Cf for high exchange (tracks of cm−2 kBq−1 m3 h−1 ) 1.59 ± 0.15 0.99 ± 0.08


022002-4

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014) TABLE III. The minimum detectable activity concentrations (Bq m−3 ) (CI = 0.95) for one month measurements using the CF and SF monitors.

TABLE II. Main specifications of SF monitor. Detector position Material Filter Dimensions

2x1 chip, fixed in pot section CF Conductive plastics Air gap, conductive sponge filter 120 mm × 70 mm × 30 mm Cf for low exchange (tracks of cm−2 kBq−1 m3 h−1 ) 1.46 ± 0.13 0.02 ± 0.001 Cf for high exchange (tracks of cm−2 kBq−1 m3 h−1 ) 1.54 ± 0.15 0.7 ± 0.06



calibration line, the monitors are exposed in three different integrated activity concentration levels of radon and thoron (about 100, 500, and 1000 kBq h m−3 ).22 These levels correspond to average activity concentrations of 140, 700, and 1400 Bq m−3 for a one-month exposure period (720 h). During the calibration, the air temperature and the humidity are examined and adjusted, while aerosol concentration is not. Typically, the temperatures are 21◦ C and 32◦ C and the relative humidities are 58% and 35% for radon and thoron, respectively. The ingoing air is filtered to inhibit the entrance of aerosol particles, and progeny of radon and thoron.37 4. Minimum detectable activity concentration

The minimum detectable activity concentration (MDAC) is the lowest activity concentration considering a certain confidence level that indicates a real measurement value eliminating the statistical variation of the background.38 According to the “Currie equation”, the MDAC is affected by the uncertainty of the background only, assuming a normal distribution for individual measurement values.39 To acquire the MDAC, the calculation of the critical detection activity concentration limit (CDACL) (also known as the decision threshold) is required using the next equation CDACL = k1−α ×

√

2UBckg .

(3)

Here α -is a certain fraction of a normalized Gaussian distribution, 0.05 for present case; 1−α -the confidence level, 0.95 for present case; k-number of standard deviation; k1 − α =1.645 for the confidence interval (CI) of 0.95; and UBckg uncertainty of the background activity concentration (Bq m−3 ). Using CDACL, the final equation for the MDAC calculation is the next MDAC = k 2 + 2CDACL = 2.706 + 4.653UBckg .

URn =

(4)

Radon MDACRn = 0.003CT n + 5.4; R 2 = 0.99 Thoron CRn ≤1000 Bq m−3 CRn >1000 Bq m−3

MDACT n = 0.089CRn + 15; R 2 = 0.89 MDACT n = 0.040CRn + 82; R 2 = 0.92 SF

Radon MDACRn = 0.006CT n + 5.1; R 2 = 0.99 Thoron CRn ≤1000 Bq m−3 CRn >1000 Bq m−3

MDACT n = 0.097CRn + 23; R 2 = 0.88 MDACT n = 0.052CRn + 130; R 2 = 0.91

5. Uncertainty of the radon and thoron measurement results

The equations for the uncertainty calculation of the radon and thoron results according to the ISO 11665-4 standard40 are given by Eqs. (5) and (6). Detailed explanation about this calculation can be found in the report of the International Standard Organization.41 With the assumption that the uncertainties of the exposure time and the scanned area of the detector surface are negligible, the UBckg of radon-thoron discriminative monitors is affected by three uncertainties: (1) of the calibration factor, (2) of the background track density, and (3) of track density that originates from the simultaneously measured radon and thoron concentrations. Turning this theory into practice, it can be seen that the elevated radon concentration level has elevated uncertainty in its absolute value that causes an increment for MDAC of thoron and vice versa. However, the increment for MDAC of radon, originating from thoron presence, is much lower because the thoron penetration efficiency (calibration factor) is lower compared to that of radon for both low and high exchange rate chambers (see Tables I and II). Therefore, the number of tracks counted on the detector surface is lower, originating from thoron, even though the concentrations of radon and thoron are the same in the monitored room. To observe the calculated MDAC of thoron for CF and SF monitors at different radon concentration levels as measured during this study, empirical linear equations can be identified to estimate the MDAC of thoron. These equations, distinguishing the background radon levels below and above 1000 Bq m−3 , are listed in Table III for CF and SF monitors. It is possible to identify empirical linear equations for MDAC of radon also; however, the gradients of these linear equations are very low compared to the gradients of the linear equations of the MDAC of thoron, expressing that the simultaneously detected thoron concentration has a weak influence on the MDAC of radon as explained above

w1 2 UT DL 2 + UB 2 − 2w1 w2 UB 2 + w2 2 UT DH 2 + UB 2 + (T DL − B)2 Uw1 2 + (−T DH + B)2 Uw2 2

(5)


Kávási et al.

022002-5

Rev. Sci. Instrum. 85, 022002 (2014)

with w1 =

and CfT n2 ; w2 βt

Uw1 2 =

=

CfT n1 ;β βt

= CfRn1 CfT n2 − CfRn2 CfT n1 ;

1 [(β − CfRn2 CfRn1 )2 UCf T n2 2 + CfT n2 4 CfRn1 2 β 4t 2 + CfT n1 2 CfT n2 2 UCf Rn2 2 +CfRn2 2 CfT n2 2 UCf T n1 2 ]

UT n =

w3 =

CfRn1 ; w4 βt

=

CfRn2 ; βt

1 [(β − CfRn1 CfT n1 )2 UCf Rn1 2 + CfRn1 4 CfT n2 2 β 4t 2 + CfRn1 2 CfT n1 2 UCf Rn2 2 +CfRn1 2 CfRn2 2 UCf T n1 2 ]

and Uw4 2 =

1 β 4t 2

[(β − CfRn2 CfT n1 )2 UCf Rn2 2 + CfRn2 4 CfT n1 2

+ CfRn2 2 CfT n2 2 UCf Rn1 2 + CfRn1 2 CfRn2 2 UCf T n2 2 ],

UT n =

+ CfT n1 2 CfT n2 2 UCf Rn1 2 + CfRn1 2 CfT n1 2 UCf T n2 2 ].

(6)

where URn and UTn are the respective uncertainties of radon and thoron measurement results, UTDH is the uncertainty of track density of the high exchange rate chamber, UTDL is the uncertainty of the track density of the low exchange rate chamber and UB is the uncertainty of the background track density. The uncertainty of the exposure time is neglected. As seen from Eqs. (5) and (6), the changes of URn and UTn are dependent on the track density of the high (TDH ) and low exchange rate chamber (TDL ) and their uncertainties (UTDH , UTDL ). Other parameters, calibration factors for low exchange (CfRn1 , CfTn1 ), for high exchange (CfRn2 , CfTn2 ), background (B), and their uncertainty (UCf Rn1 ), (UCf Tn1 ), (UCf Rn2 ), (UCf Tn2 ), (UB ) are constant during the calculation process. For example, using the calibration factor data from Table I and t = 720 h (one month), Eqs. (5) and (6) can be expressed as

(7)

2.0 × 10−6 UT DH 2 + 1.9 × 10−6 UT DL 2 + (T DH − B)2 8.8 × 10−16 + (−T DL − B)2 2.0 × 10−15 .

(8)

Neglecting the parts in each equation where the multiple constants are negligible (about 10−11 , 10−17 ) relative to the multiple constant ranges (10−6 , 10−7 ) of other parts gives: (9) URn = 7.2 × 10−7 UT DL 2 − 7.1 × 10−7 UB 2 ) and UT n =

[(β − CfRn1 CfT n2 )2 UCf T n1 2 + CfT n2 4 CfRn2 2

7.2 × 10−7 UT DL 2 − 7.1 × 10−7 UB 2 + 7.4 × 10−11 UT DH 2 + (T DL − B)2 6.3 × 10−17 + (−T DH − B)2 3.7 × 10−17

URn =

and

1 β 4t 2

w3 2 UT DH 2 + UB 2 − 2w3 w4 UB 2 + w4 2 UT DL 2 + UB 2 + (T DH − B)2 Uw3 2 + (−T DL + B)2 Uw4 2

with

Uw3 2 =

Uw2 2 =

2.0 × 10−6 UT DH 2 + 1.9 × 10−6 UT DL 2 .

(10)

For the passive discriminative monitors, the uncertainty of the radon measurement (URn ) depends on the uncertainty of the track density of the low exchange rate chamber (UTDL ) and on the uncertainty of the background track density (UB ). However, in an indoor environment the influence of the UB is weak as the UTDL is some orders of magnitude higher than UB . The result of this phenomenon can be seen in Fig. 3, where the relative standard deviation of the radon measurement result is decreased by the elevated radon concentration level (increased track density results in lower uncertainty), while


022002-6

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

FIG. 3. Relative standard deviation of radon results of the CF monitor at different radon concentration levels versus thoron concentration.

slight influence from the thoron concentration level can be detected at lower radon concentration (hundred) and no influence at higher levels (thousand). The uncertainty of the thoron measurement (UTn ) depends on the uncertainty of the track density of the high (UTDH ) and low (UTDL ) exchange rate chamber. Since these values are commensurable, both radon and thoron concentrations have an influence on the UTn as it can be seen in Fig. 4. The relative standard deviation of the thoron measurement result is decreased by the elevated thoron concentration level (increased track density results lower uncertainty), while influence from the elevated radon concentration level also can be detected (slopes of the lines are higher than zero).

Comparing Eqs. (9) and (10), it has to be recognized that the UTn is typically higher compared to the URn as the multiple constants for UTDH and UTDL are higher in Eq. (10) than in Eq. (9). This phenomenon can be also clearly understood from Figs. 3 and 4. For observed radon and thoron levels about 200 and 1000 Bq m−3 , the relative standard deviations are about 10% and 3% for radon in a narrow range, while they are more than 20% (20%–45%) and 6% (6–20%) for thoron in a wide range. These differences originate primarily from the different detection responses of radon and thoron. The monitors always have a lower response from thoron as it has weaker diffusion ability through the diffusion chambers due to its shorter halflife compared with radon. This is why the calibration factor is lower for thoron than radon (see Tables I and II). Standardized

FIG. 4. Relative standard deviation of thoron results of the CF monitor at different thoron concentration levels versus radon concentration.


022002-7

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

FIG. 5. Minimum detectable thoron activity concentration for CF and SF monitors versus parallelly detected radon concentrations.

for radon, the thoron detection responses are lower with 48% and 55% values for CF and SF monitors, respectively.

6. Comparison of the measurement results

The significant difference between CF and SF results is checked, taking into account the measured results and their uncertainties. In this process, the absolute difference (AD) of the measured values is compared to the uncertainty of the absolute difference (UAD ).42 The absolute difference is calculated as AD = |CCF − CSF |,

(11)

while the uncertainty of the absolute difference is calculated as (12) UAD = UCF 2 + USF 2 , where CCF and CSF are radon and thoron values measured with CF and SF monitors and UCF and USF are uncertainties of radon and thoron values measured with CF and SF monitors. If AD ≤ 1.96UAD , then CCF and CSF are accepted as equal values, corresponding to a confidence interval of 95%. As the presence of radon influences, the MDAC for thoron and the probability of elevated thoron concentration level is low, in many cases thoron concentration under the MDAC is expected. Due to the different parameters of the CF and SF diffusion chambers, their MDACs are different. The empirical equations (CRn ≤1000 Bq m−3 ) from Table III are used and the estimated MDAC for thoron versus measured radon concentration is obtained and plotted in Fig. 5. In this figure, three different cases can be distinguished basically in comparing SF and CF thoron results. Case A: The detected thoron concentrations are higher than the MDAC for both monitor types. Case B: The detected thoron concentrations are higher than the MDAC for CF but lower than the

MDAC for SF. Case C: The detected thoron concentrations are lower than the MDAC for both monitor types. For case A, the comparability is evident. For case B, no significant difference between CF and SF values is apparent when the SF MDAC is used for the comparison and for the correlation analysis. For case C, no significant difference is also apparent when the CF and SF MDACs are used for the comparison and for the correlation analysis. Considering the relatively high MDAC of thoron in Fig. 5 relative to the MDAC of radon which is about 10 Bq m−3 at 1000 Bq m−3 thoron concentration, the presence of thoron (although in a very low concentration) is not excluded; therefore, the validity of using the MDAC value in the correlation analysis can be confirmed.

B. Measuring sites

Measurements with CF and SF radon-thoron discriminative monitors in a manganese mine (MM) and in the Lake Cave of Tapolca (LC) and Hospital Cave of Tapolca (HC) were made from December 2006 until November 2007; and in a bauxite mine (BM) from February 2007 until January 2008. Measurements were carried out at eight points in the BM, ten points in the MM, and three points each in the LC and HC. CF and SF monitors were fixed to the walls at the same points, directly next to each other and were replaced every month. Sampling aerosol concentration measurements in summer, wind speed measurement in winter and summer and humidity and temperature measurements in every month were made at the measurement points. From hydrological and geological aspects the Lake Cave of Tapolca and the Hospital Cave of Tapolca are in the same cave system, but since there is no passable connection between the two, the difference in names is appropriate. The measuring sites are shown on the map in Fig. 6.43 The properties of the sites and a list of the measurement points are given in Tables IV and V, respectively.


022002-8

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

FIG. 6. Locations of the measurement sites in Hungary.

III. RESULTS AND DISCUSSION A. Results of radon measurements

Average radon concentration values measured at the different measurement sites are given as a monthly spread in Fig. 7.

The highest radon concentration levels (about 10 000 Bq m−3 ) were found in the LC and HC. This value was characteristically a few thousand Bq m−3 at the MM, while it was a few hundred Bq m−3 at the BM. As indicated in Table IV, mechanical ventilation was only applied in the mines. Ventilation in MM was used for 40 h per week (24%

TABLE IV. Main properties of the measuring sites BM

MM

HC and LC

Aver. air temp. (◦ C)

12 (main air current) 15 (partial ventilation) 12 and 21 14 (partial ventilation) 13 (main air current) Aver. rel. humidity (%) 100 100 100 and 100 Aver. depth from surface (m) 120 200 13 and 18 Product Bauxite Mn-carbonate ore 85% NAa with boehmite and gibbsite Mn-oxide ore 15% Typical mining waste Triassic dolomite, Liassic clayey Pliocene limestone, Eocene limestone, marlstone, Upper Triassic clayey marlstone, limestone, dolomite, clayey-coaly beds black shale clay 11 7 NA Ave. area of galleries (m2 ) 158 and 2.4 1000 and 2.4 NA Ave. achiev. of vent. (m3 min−1 ) & wind speed (m s−1 ) Aver. aerosol conc. 46 600 59 300 560 and 130 (13 000–152 000) (7 500–160 000) (400–710) and (30–260) (min-max) (particles cm3 ) Typical power source of diesel pneumatic NA mining machinery Number of underground 121 50 3 and 13 workers Number of shifts 3 on weekdays 1 on weekdays 1 on weekdays (no work on weekends (no work on weekends and Saturday (HC) and holidays) and holidays) 1 on weekdays, weekends and holidays (LC) a

NA-not applicable.


022002-9

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

TABLE V. Measuring points at the different sites.

No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8

BM Inclined shaft (Conveyor shaft) Airway of deposit No. 40 Warming room No. 5 Near sump No. 2 Near air dam of deposit No. 40 Exploitation site, deposit No. 40 Exploratory site in 1st horizon of deposit No. 2 Exploitation site, deposit No. 2

No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9 No. 10

MM Exploitation site, Level 178 maBSa Exploitation site, Level 186 maBS Exploitation site, Level 186 maBS Exploitation site, Level 186 maBS Exploitation site, Level 180 maBS Water treatment center Level 175 maBS Shaft No. 3, Level 175 maBS Exploitation site, Level 192 maBS Exploitation site, Level 203 maBS Out-coming air current Level 241 maBS

No. 1 No. 2 No. 3

Six-bed hall Large hall Small chamber

No. 1 No. 2 No. 3

LC Batsanyi hall Loczy hall Exit of the south passage

HC

a

maBS-metre above Baltic Sea.

of the measurement period), while in BM it was 120 h per week (71% of the measurement period). The rest of the time, the radon concentration of the mine galleries would only be changed by natural air movement. In the by-galleries (partially ventilated galleries), poor natural air change was a char-

acteristic causing considerable accumulation of radon. Integrated results of the track detectors also included the radon concentration values of periods without mechanical ventilation. Therefore, the lower radon concentration observed in the bauxite mine was evidently due to the longer period of mechanical ventilation. In the caves, where artificial ventilation was not applied the efficiency of natural air change was much lower; therefore, radon concentration elevation was expected. Table VI summarizes previous radon and thoron measurement results and Table VII summarizes the present radon measurement results. Comparing the results of the present study to the results of the previous measurements, good agreement was found among MM and BM results. In the LC, the annual arithmetic mean of the measurements of the previous 4 years changed in a wide range, between 6411 and 9726 Bq m−3 . Compared to this, the results of the CF and SF monitors seemed to be appropriate. However, in the HC the annual arithmetic mean of the measurements of the previous 4 years changed in a narrower range, between 4320 and 5160 Bq m−3 . Therefore, the present data in the HC could be considered as slightly low for both types of monitors. For the HC, comparison of the GM average values was doubtful, as the seasonal radon concentration changes in the HC should be taken into consideration, since there was a maximum level during warm summer months and a minimum level during cold winter months (humid continental climate). Therefore, the measurements performed using CF monitors previously from May to October could not be compared to the whole year GM average. Previous studies have already proved seasonal changes of radon in the HC and MM. The great differences between summer and winter data can be explained by the chimney effect: the volume of the air change is affected by the change of the outdoor air temperature (pressure). The temperatures in the mines and caves are rather constant, changing only within narrow limits. When the outdoor temperature falls below the temperature of the air in the cave or mine, the warmer air with smaller density in the galleries flows outwards, while the

FIG. 7. Monthly average radon concentration at the different workplaces.


022002-10

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

TABLE VI. Radon and thoron results at the sites from earlier surveys.

AMa (GM)b

Radon (Bq m−3 ) Min-max (GSD)c

BM

... (275)

MM

Site

HC

LC

AM (GM)

Thoron (Bq m−3 ) Min-max (GSD)

Below the MDAC (%)

Measurement period

... (1.80)

... (341)

... (2.59)

... 0

... (839) 817 ...

... (1.62) 575–997 ...

... (261) ... ...

... (2.04) ... ...

... (2763) 4630 (2620)

... (2.11) 510–12 400 (2.63)

... (751) ... ...

... (4489) 8287 ...

... (1.72) 1154–15033 ...

... (1028) ... ...

Device

Reference

6 months (May–Oct)

CF

34

... 36 ... ...

6 months (May–Oct) 1 year

CF

34

NRPB/CR-39

3

... (2.56) ... ...

... 17 ... ...

6 Months (May–Oct) 3 years 1 year

CF

34

DATAQUA Rn monitor NRPB/CR-39

45 46

... (2.38) ... ...

... 31 ... ...

6 months (May–Oct) 4 years

CF

34

NRPB/CR-39

47

a

AM-arithmetic mean. GM-geometric mean. c GSD-geometric standard deviation. b

outdoor colder and radon-poor air with greater density flows inwards, reducing the radon concentration in the galleries. In the summer, this process is reversed causing reduced ventilation with higher radon level. This air change is more intensive if the air from the passage can exit at an outgoing point located at a higher level (lower pressure point) related to the ingoing point with a given height (higher pressure point).44 This is why the outgoing point of the exiting air current is always located higher than the other shafts in mines. In the BM due to the long period of ventilation, seasonal changes of radon were not clearly observable. The seasonal trend was also missing in the LC, highlighting inefficient ventilation, which was also confirmed by the lowest aerosol concentration with 130 particles cm−3 emphasizing the absence of incoming aerosol contamination (Table IV). In Table VIII, results of CF and SF monitors are compared using linear regression analysis. The comparison was performed on all measurement data of the given measurement area, distinguished as radon levels above and below TABLE VII. Radon results of the present study at the different measuring sites. Site

Mon.

AMa (SD)b

GMc (GSD)d

BM

SF CF SF CF SF CF SF CF

550(497) 350(321) 1258(788) 887(604) 3403(3075) 2510(2341) 8512(1955) 6239(2057)

433(1.89) 265(1.86) 1096(1.7) 755(1.82) 2225(2.57) 1521(2.95) 8331(1.23) 5900(1.42)

MM HC LC a

AM-arithmetic mean. SD-standard deviation. c GM-geometric mean. d GSD-geometric standard deviation. b

Min-max 52–4033 6–3110 91–3813 50–2517 301–9545 154–7913 5184–13661 1841–9663

1400 Bq m−3 since this value was the maximum monthly exposition radon level for which calibration factors have been given. As seen in Table VII, different average annual radon concentration levels were measured at the different workplaces. This phenomenon allowed for the monitors to be compared at three different radon levels: low at 500 Bq m−3 (BM), intermediate at 1000 Bq m−3 (MM), and high at 5000 Bq m−3 (HC, LC). The aerosol concentrations of two levels were differentiated within the range of 20–1000 nm: a rather low concentration of a few hundred particles cm−3 in the caves, and a rather high concentration of several tens of thousands of particles cm−3 in the mines (compared to a few thousand particles cm−3 in average buildings and outdoors). No difference could be attributed to the relative humidity as it was characteristically about 100% in all measurement areas (Table IV). Since the HC and LC have the same hydrological and geological environments, they were analyzed together, however, different air change efficiency was observed at these areas. The results of SF monitors were higher than those of the CF monitors at each measurement location, as seen in Fig. 7. The data of Table VIII also confirmed this as the variable coefficient (slope) for the CF data was higher than 1 for each case where the radon values followed a normal distribution. When all measured data were examined, the largest difference, between the data of CF and SF, was seen at the low radon level (BM), meaning a variable coefficient of 1.36 ± 0.06. This value was 1.11 ± 0.06 and 1.16 ± 0.06 for intermediate and high levels, respectively. This meant that, the SF monitor data were higher by 36(±6)%, 11(±6)%, and 16(±6)% at the different radon levels relative to the CF monitor data. The difference further increased when taking the intercept coefficient into consideration (74, 270, and 880), resulting in a minimum 15% further increment (15(±7)%, for low level (500 Bq m−3 ); 27(±7)% for intermediate level


022002-11

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

TABLE VIII. Regression analysis results on radon data of CF and SF monitors (CI = 0.95).

Equation

R2a

Sum 1400< 1400 ≥

yf = 1.36(±0.06)xg + 74.4(±36) y = 1.11(±0.60)x + 980(±946) y = 1.25(±0.07)x + 90.3(±27)

0.864 2 0.76

Sum 1400< 1400≥

y = 1.11(±0.06)x + 270(±67) y = 0.43(±0.13)x + 1426(±185) y = 1.04(±0.07)x + 160(±41)

0.728 0.207 0.774

Sum 1400< 1400≥

y = 1.16(±0.06)x + 880(±326) y = 1.02(±0.09)x + 1769(±518) y = 0.56(±0.21)x + 540(±124)

0.831 0.701 0.368

Rn

Fb BM 459 3 240 MM 316 12 227 HC and LC 344 131 6.9

Dfc

Ndd

Nsde (%)

75 2 71

Yes No Yes

17.7 1.0 16.7

118 48 68

Yes No Yes

19.2 1.7 17.5

70 56 12

Yes Yes No

8.3 8.3 0

a

Determination coefficient. F statistic. Degree of freedom. d Normal distribution. e No significant difference. f SF data. g CF data. b c

(1000 Bq m−3 ); 18(±6)% for high level (5000 Bq m−3 )). Accordingly, the SF monitor data were higher by 51(±9)% for low level, 38(±9)% for intermediate level, and 34(±9)% for high level, relative to the CF monitor data. Examining these items for values below 1400 Bq m−3 (BM, MM) and over 1400 Bq m−3 (HC, LC), 45(±9)%, 20(±8)%, and 37(±14)% were the differences. In the case of the elevated radon level, the difference should have decreased, due to the statistical characteristics of radioactive decay, causing more accurate results for higher radon concentration activity levels. This trend was not confirmed as it was overcome by the influence of the environmental parameters. Radon data sensitivity of CF and SF to environmental parameters has been tested previously in two different studies. In the first one that was carried out by the Radon Laboratory Group at the Institute of Energetic Techniques of the Technical University of Catalonia in Barcelona, Spain, the dependency on temperature (10◦ C–30◦ C) and relative humidity (30%–80%) was checked for about 650 kBqh m−3 integrated radon concentration (about 1000 Bq m−3 for one month).48 During that study no effects of temperature and humidity were confirmed. In the present case, the temperature of the calibration (21◦ C) and during the measurement (12◦ C–21◦ C) was in the ranges of the Spanish test study; thus the influence of temperature on the radon data of the present study can be excluded. In the second sensitivity test, which was carried out at NIRS, the dependency on relative humidity (32%–87%) and aerosol concentration (0–6000 particles cm−3 , mono-modal aerosol distribution −50 nm) was checked for about 500 kBqh m−3 integrated radon concentration (about 700 Bq m−3 for one month).49 Analysing the data of the NIRS study (ttest, CI = 0.95, Df = 18) no statistically strong tendency was discovered either for temperature or for aerosol concentration influence. In the present study, the relative humidity (100%) and the aerosol concentration (and size) were over the test range, thus the influence effect of these environmental parameters cannot be excluded. The considerable difference

between CF and SF monitor data in the mines could be caused by the elevated aerosol concentration decreasing the diffusion capabilities of the diffusion chambers which is more intensive for the CF than the SF monitors. Comparison of the effect of humidity could not be done as it was 100% at each place, conversely its effect on aerosol deposition rate should be considered. Technically, the same type of CR-39 detectors was used for the two different monitors, applying the same etching conditions and using the same track counting system with the same adjusted parameters. Beyond the environmental parameters, the monitor sensitivity of radon level change could be the source of the radon result deviation. As the radon and thoron calibration is carried out with stabilized radon and thoron concentration levels, there is no information about the monitor response when the radon and thoron concentrations change with different frequencies as it would be at real measurement sites. If one monitor has a slower response for radon or thoron change it will detect lower radon or thoron level compared to another which has a faster response. To confirm these assumptions, further laboratory tests are required. B. Results of thoron measurements

Fig. 8 shows the monthly average thoron concentrations measured at the different sites, while Table IX summarizes the thoron measurement results. The thoron concentration was below the MDAC in about 50% of the measurements in the MM, HC, and LC. In April, the CF monitors detected no thoron concentration at any measurement points in the MM. This did not happen in the BM; however, it happened several times in the caves. In the HC, the CF monitors did not provide any reliable thoron concentration for four months (January, February, March, December), and neither did the SF monitors for four months (February, March, May, June). In the LC, the CF monitors did not provide reliable thoron concentrations for three months (January, February, March), and the SF


022002-12

Kávási et al.

Rev. Sci. Instrum. 85, 022002 (2014)

FIG. 8. Monthly average thoron concentrations at the different workplaces.

monitors did not for two months (January, February). The fact that thoron level was undetectable in the HC during the winter months, when the MDAC is lower since the radon level is also lower, may be attributed to the radon concentration being significantly lowered during winter months. However, this was inconsistent with the fact that this phenomenon occurred even in June, when the radon concentration was considerable. The same inconsistency was also supported by the results of LC, where radon concentration did not change with the seasons; however, thoron fluctuated considerably as there was no thoron in March, while 13 kBq m−3 was measured in April (SF). This anomaly would be owing to extreme environmental conditions such as 100% relative humidity that prevented the monitors from providing reliable thoron data. Considering the short half-life of thoron, air mixing around the monitors can

TABLE IX. Thoron results of the present study at the different measuring sites. Site

Mon.

AM (SD)

GM (GSD)

Minmax

Mean MDAC (Min-max)

Below the MDAC(%)

BM

SF

161 (148) 154 (210)

142 (1.69) 117 (1.78)

Invited article: radon and thoron intercomparison experiments for integrated monitors at NIRS, Japan.

Evaluation of three types of passive samplers for measuring 1,3-butadiene and benzene at workplaces.

In situ characterization of mixed-wettability in a reservoir rock at subsurface conditions.

Isothiazolinones in commercial products at Danish workplaces.

In planta passive sampling devices for assessing subsurface chlorinated solvents.

Invited Review Article: Pump-probe microscopy.

Wellness at work: Building healthy workplaces.

Passive sampling of perfluorinated chemicals in water: in-situ calibration.

Invited article: Richard D. Turk Memorial Lecture--parasites' progress.

Invited review article: The Chandra X-ray Observatory.

Invited review article: Advanced light microscopy for biological space research.

Invited article: Characterization of background sources in space-based time-of-flight mass spectrometers.

Effects of dentifrices on subsurface dentin tubule occlusion: an in situ study.

Invited Review Article: Error and uncertainty in Raman thermal conductivity measurements.

Invited editorial on "Surface tension in situ in flooded alveolus unaltered by albumin".

Aging trends: Hungary : Hungary has oldest population in Eastern Europe.

Monitoring of bacterial pathogens at workplaces in power plant using biochemical and molecular methods.

Neuroscience in Hungary.

Suicide in U.S. Workplaces, 2003-2010: a comparison with non-workplace suicides.

Assessing Latrine Use in Rural India: A Cross-Sectional Study Comparing Reported Use and Passive Latrine Use Monitors.

Invited Article: Autonomous assembly of atomically perfect nanostructures using a scanning tunneling microscope.

Neuro-ophthalmology in Hungary.

HIV in Hungary.

Invited Frontiers Commentary. Tier Climbing Article: Redefining Neuromarketing as an Integrated Science of Influence.