Journal of Environmental Radioactivity 129 (2014) 57e62
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Radon dispersion modeling and dose assessment for uranium mine ventilation shaft exhausts under neutral atmospheric stability Dong Xie a, b, *, Hanqing Wang c, Kimberlee J. Kearfott d, Zehua Liu b, Shunquan Mo b a
School of Energy Science and Engineering, Central South University, Changsha 410083, China School of Urban Construction, University of South China, Hengyang 421001, China c School of Civil Engineering, University of Hunan Technology, Zhuzhou 412007, China d Department of Nuclear Engineering and Radiological Sciences, University of Michigan, Ann Arbor, MI 48109-2104, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 August 2013 Received in revised form 3 December 2013 Accepted 4 December 2013 Available online 28 December 2013
In the present study, the roles of atmospheric wind profiles in the neutral atmosphere and surface roughness parameters in a complex terrain were examined to determine their impacts on radon (222Rn) dispersion from an actual uranium mine ventilation shaft. Simulations were completed on 222Rn dispersion extending from the shaft to a vulnerable distance, near the location of an occupied farmhouse. The eight dispersion scenarios for the ventilation shaft source included four downwind velocities (0.5, 1.0, 2.0 and 4.0 m s1) and two underlying surface roughness characteristics (0.1 m and 1.0 m). 222Rn distributions and elevated pollution regions were identified. Effective dose estimation methods involving a historical weighting of wind speeds in the direction of interest coupled to the complex dispersion model were proposed. Using this approach, the radiation effects on the residents assumed to be outside at the location of the farm house 250 m downwind from the ventilation shaft outlet were computed. The maximum effective dose rate calculated for the residents at the outside of the farm house was 2.2 mSv y1, which is less than the low limit action level of 3e10 mSv y1 recommended by the International Commission on Radiological Protection (ICRP) occupational exposure action level for radon. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: 222 Rn release Atmospheric dispersion Numerical modeling Radiation dose evaluation
1. Introduction Radon (222Rn) emitted from uranium mine ventilation shaft exhausts could constitute a major source of environment contamination and consequently a potential health hazard to the nearby population. Due to their alpha-emitting short lived progeny 218Po and 214Po, 222Rn have long been recognized as main causative agent for lung cancer when presented in high radon inhalation, such as those encountered in uranium mining areas (Evans et al., 1981). The geographical features of the dispersion region and the meteorological conditions are important for evaluating the dispersion of 222 Rn from uranium mine shaft exhausts. Essential parameters to be considered include the 222Rn emission concentration as it leaves the shaft, gas emission velocity, shaft location and height, trees and topography, wind speed and direction, atmospheric stability (Bruce and Werner, 1990) and precipitation.
* Corresponding author. School of Urban Construction, University of South China, Hengyang 421001, China. Tel.: þ86 734 828 2512; fax: þ86 734 828 2312. E-mail address: [email protected] (D. Xie). 0265-931X/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvrad.2013.12.003
Full-scale field measurements, wind tunnel experiments, and computational fluid dynamics (CFD) models have been used to study pollutant dispersion for complex underlying surface conditions. For some field measurement situations, it is hard to simultaneously control operative and intertwined parameters effects (Ana Pilar et al., 2002) such as atmospheric conditions (wind speed, wind direction), topography and geography (underlying surface roughness, mountain height and shaft height, width, shape). The need to utilize hazardous radioactive sources and also the difficulties in creating appropriate boundary conditions similarity may limit the efficiency of wind tunnel experiments (Sharma et al., 2005). However, CFD works well for this situation. CFD has been proved to be a very powerful and efficient tool for the studies of radionuclides dispersion with the factors considered individually or in combination with the wind field effect (de Sampaio et al., 2008). Several previous studies have involved simulations of the atmospheric dispersion of nuclear power plant (NPP) emissions (Srinivas and Venkatesan, 2005; Basit et al., 2008), including from the Chernobyl (Hiroaki and Masamichi, 2008) and Fukushima accidents (Leelössy et al., 2011). 222Rn concentration dispersion and the effective dose evaluation obtained in this study differ from the NPP work not only due to the different radionuclide, but because the
58
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
release was continuously from a low height and the topographical and geographical features and surrounding the release are very different, namely in mountainous rural terrain. Little information was available in the literature about the numerical analysis of 222Rn dispersion and dose evaluation of 222Rn from uranium mine ventilation shafts. Some CFD modeling of the situation has been finished, but it was limited to a single location. This work did, however, included field measurements that provided solid validation for CFD modeling of this problem (Dong et al., 2012). In current study, 222Rn dispersion under neutral atmospheric stability conditions was analyzed using three-dimensional CFD simulation for a specific area above a uranium mine in one province of China. This particular location included an occupied home and farmed area around it. CFD modeling was accomplished using Fluent (Ansys Fluent, 2010), a commercially available and widely used tool incorporating several turbulence models (ANSYS Fluent 13.0.0, ANSYS Inc., Southpointe, 275 Technology Drive, Canonsburg, PA 15317 USA). 222Rn distributions and elevated pollution regions in the surrounding area of uranium mine ventilation shaft were calculated. For this case, public effective dose rate calculation methods were developed and radiation dose rate to the farmers living and working in the vicinity of a house 250 m downwind from the shaft was evaluated. 2. Models and computational methods 2.1. Geometric model The geometric model considered in this study was based upon an actual ventilation shaft located in uranium-bearing mountains in China. Discrete data points for the model were extracted from an elevation contour map and processed with specialized Fluent preprocessing software, Gambit (Ansys Fluent, 2010). The overall computational dimensions of mountain were 400.0 m (L) 400.0 m (W) 150.0 m (H) in the X, Y and Z directions, respectively. The ventilation shaft outlet was modeled to match its actual dimensions of 2.7 m 2.7 m, extending 2.0 m above the ground’s surface. In the present study, Temperature of the exhaust is set to be the same as the atmosphere because of neutral atmospheric stability and large amount of exhaust air rate. 222Rn was emitted from uranium mine ventilation shaft at a certain flow rate of 3.0 m s1, then mixed with air and dispersed in three directions into atmosphere as shown in Fig. 1a. 2.2. Mathematical models Considering that the atmospheric temperature could be regarded as homogeneous with heights in the computation region under the neutral atmospheric stability, CFD simulation was based on the governing equations of continuity, momentum, pollutant transport, turbulent energy and turbulent dissipation rate, as shown in Table 1. Xiaomin et al. (2005) investigated the influence of complex geometry on pollutant dispersion comparing different keε models with wind tunnel measured data for optimization of turbulence models. Their comparison results show that the standard keε model provides the best simulation results (Brian and Dudley, 1974), while Renormalization Group （RNG） keε turbulence models based on RNG theory (Victor and Steven, 1986) and modified keε turbulence models (Chen and Kim, 1987) over-predict the pollutant concentrations. Standard keε closure was thus chosen as the turbulence model in this study. The parameters of equations (1)e(6) depicted in Table 1 are defined as: X is the coordinate axis in the direction i (i ¼ 0, 1, 2), ui corresponds to the mean velocity in i direction, r is the air density, t is time, P is pressure, mt is the turbulent viscosity, m is the molecular viscosity, and g is the
Fig. 1. Fig. 1a. Map of surface above uranium mine showing actually topographic details, including the location of the shaft and the house (250 m downwind), along with the computational slice X ¼ 0 and the southerly wind direction perpendicular to the Y ¼ 0 plane. Fig. 1b. Unstructured computational grid used for finite volume analysis for the CFD model of the actual topographic surface.
gravitational acceleration. K is the turbulence kinetic energy, and ε is the dissipation rate of turbulence kinetic energy. Cm, C1ε, C2ε, sk, and sε are empirical and experimental constants fixed as 0.09, 1.44, 1.92, 1.0, and 1.3, respectively (Brian and Dudley, 1972). C is the 222 Rn average concentration in the air, u is the velocity vector of the 222 Rn, D is the effective diffusion coefficient of 222Rn in air assigned a value of 5 103 m2 s1, which is from the work of Guo (Qiuju et al., 1995), l is the radon decay constant equal to 2.1 106 s1. 2.3. Numerical codes and solution methods CFD modeling was conducted with the code Fluent in this paper. The simulation involved the finite volume discretization of the equations of motion, a geometrical model consisting of an unstructured grid volume made of tetrahedral cells, various matrixinversions routines, and the keε turbulence model (Jie et al., 2009). The coupling between velocity and pressure was accomplished through the SIMPLE algorithm (Patankar, 1980). The central differencing scheme was used in diffusion term and advection term, while an upwind differencing scheme was used in source term. In such a complex terrain area, the unstructured grid system is the most efficient for CFD simulations (Hong et al., 2005), so this approach was selected. The computational domain was built using tetrahedral cells with a finer resolution nearest the ventilation shaft outlet and the mountain ground surface. In the CFD model, a nonuniform tetrahedral grid of approximately 370,000 cells was
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
59
Table 1 Governing equations. Continuity equation vðrui Þ=vXi ¼ 0
(1)
Turbulent momentum equation v rui uj =vXi ¼ vP=vXj þ v ðmt þ mÞ vui =vXj þ vuj =vXi vXi þ rg
(2)
Standard turbulence kinetic energy equation vðrui kÞ=vXi ¼ v½ðm þ mt =sk Þðvk=vXi Þ=vXi þ mt vuj =vXi vui =vXj þ vuj =vXi rε
(3)
Standard dissipation rate of turbulence kinetic energy vðrui εÞ=vXi ¼ v½ðm þ mt =sε Þðvk=vXi Þ=vXi þ C1ε mt ðε=kÞ vuj =vXi vui =vXj þ vuj =vXi rC2ε ε2 =k
mt ¼ Cm r k2 =ε
(4)
(5)
Radon transport equation vðCuÞ=vXi ¼ D v2 C=vXi2 þ Q lC
(6)
created. The triangular mesh of the computational domain of the topographic surface was shown in Fig. 1b. 2.4. Boundary conditions and parameters Inlet boundary condition was used for the atmospheric wind with specified velocity and direction. Simulations were performed for atmospheric wind velocities of 0.5, 1.0, 2.0 and 4.0 m s1 for class D, or neutral, atmosphere stability condition. For the location of the ventilation shaft, neutral atmosphere stability condition occurs 52.1% of the time according to local meteorological conditions. The probabilities of atmospheric wind speeds of 0.5,1.0, 2.0, 4.0 m s1 are 8%, 17%, 61%, 9% and 5%, respectively. The remaining 5% is the wind speed above 4.0 m s1 and its influence is neglected in this study. A user-defined subroutine compiled by Cþþ was developed and used in the analysis for boundary conditions, for example, power law velocity profiles applied at atmospheric wind inlet and turbulence kinetic energy and dissipation rate applied at shaft outlet with the Fluent code. The turbulence kinematic energy k and dissipation rate ε at the shaft outlet were defined in Table 2. The parameters of equations (7)e(9) depicted in Table 2 are defined as: uY is the wind speed at height Z, u0 is the average wind speed at the height of a reference height Z0 (Z0 ¼ 10 m), and the exponent a is a function of both the atmosphere stability in the layer and the underlying surface characteristics. u is the average wind velocity, I is fractional turbulent component, and L is turbulence integration scale. Table 2 Boundary conditions and calculation parameters settings. Wind inlet uY =u0 ¼ ðZ=Z0 Þa ; Z0 ¼ 10m
(7)
z ¼ 0.1 m, a ¼ 0.19 z ¼ 1.0 m, a ¼ 0.25 (Oliver, 1953)
Shaft outlet k ¼ 1:5ðu IÞ2 ε ¼ Cm0:75 k1:5 =l;
l ¼ 4ð0:09kÞ0:5 Z00:25 Z 0:75 =U0
Side and top Wall Surface roughness
(8)
ðHong et al:; 2005Þ Outflow Non-equilibrium wall function z ¼ 0.1 m, 1.0 m
(9)
Another inlet boundary condition was the radon-bearing exhaust jet plume from the shaft outlet. The shaft exhauss velocity was set to its average measured value of 3.0 m s1. 222Rn concentration was measured at the side of the top of the ventilation shaft to be 8000 Bq m3 and the value was used as the source term. For the outlets at the end, the top and the lateral sides of the domain, outflow boundary conditions were assumed. The ground surface and the lateral sides of the shaft outlet were treated as walls with no-slip velocity boundary condition, and a non-equilibrium wall function was applied for near wall treatment (Jie et al., 2009). The wind blowed in the southerly direction towards the farmer’s house occurs in 25% of the time. Based upon local topographical and geomorphologic features across the mountain being modeled, as well as the effects of seasons for spring-summer and autumnwinter, two underlying surface roughness (0.1 m, 1.0 m) were considered in the study. In this study, surface roughness of 1.0 m represented the seasons effects of spring-summer as there were many vegetations like leaves on trees or grass on the ground, which would pose relative higher underlying surface frictional resistance. The surface roughness of 0.1 m represented the less vegetation effects of seasons for autumn-winter. 2.5. Radiological dose calculation methods Using the radon concentration values obtained from CFD calculation, the effective dose rate for the residents assumed to be outside at the location of the farm house was calculated with standardized methods (UNSCEAR, 2000). The corresponding dose conversion factor value, i.e. the effective dose rate received by adults per unit 222Rn activity per unit of air volume, used for the computation, 9.0 106 mSv per Bq m3 h, was remained constant for several years (UNSCEAR, 2000a, 2009b). The effective dose rate to the public from radon at different distances from ventilation shaft outlet was estimated using following equation:
E=T ¼ C F D
(10)
where, E/T was public personal effective dose rate (mSv h1), C was radon concentration (Bq m3), F was an equilibrium factor applicable to outdoors and it was 0.6 found from UNSCEAR report (UNSCEAR, 2000), T was the annual exposing hours obtained from
60
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
the questionnaires finished by the occupants who lived and worked in the environment of interest (h y1), and D was the dose conversion factor (9.0 106 mSv per Bq m3 h). Considering that atmospheric wind speeds probability distribution, the average effective dose rate posed by 222Rn could be depicted as:
ETotal =T ¼
X
fV CV F D
(11)
V
where, ETotal/T was the sum of the public personal effective dose rate per hour (mSv h1), CV was radon concentration under atmospheric wind speeds V of 0.5, 1.0, 2.0, 4.0 m s1 (Bq m3), fV was the fractional probability of a given atmospheric wind speed. 3. Results and discussion 3.1. Radon concentration distribution For the present simulation, the exhaust emissions went downwind and the exhaust emissions would be trapped in the valley which located at the distance of 120 m from shaft outlet exhausts. The effect was caused by that the shaft height was relatively low at a certain shaft exhausts outlet rate and the location of the shaft was located on the hillsides of mountains. Due to the valley accumulation effects, 222Rn concentration with the region of 150 m on the downstream direction were high and the plume cannot transport efficiently, as shown in Fig. 2. When ambient wind speed was 0.5 m s1, the diffusion ability was not significant. 222Rn concentration from shaft outlet exhausts was higher than 500 Bq m3 within the region of 100 m. When atmospheric wind speed increased from 0.5 m s1 to 1.0 m s1, 222 Rn concentration at surface roughness of 0.1 m dropped from 500e378 Bq m3, while 222Rn concentration at surface roughness of 1.0 m dropped from 556e390 Bq m3 at the distance of 50 m. As shown in Fig. 2, ambient wind profiles played the dominant effect in radon dispersion process. Besides wind direction, wind speed was another important factor as higher wind speeds leaded to more effective pollutants dilution (Hong et al., 2000). Fig. 2 also indicates that the effect of surface roughness was negligible when the atmospheric wind speeds exceeded 2.0 m s1. The radioactive pollution along downwind direction on local region over 200 m was within the limits of 100 Bq m3, as the dispersion ability of 222Rn was strong and 222Rn could travel further at higher atmospheric wind speeds. Compared with predicted downstream radon concentrations for an assumed surface roughness of 0.1 m, 222Rn concentrations for simulated surface roughness of 1.0 m were larger, as shown in Fig. 3. An increased number of plants and trees, associated with a high surface roughness, were expected to cause the accumulation of 222 Rn. As shown at about 120 m from the exhaust outlet in the downwind direction, radon concentrations at surface roughness of 1.0 m were remarkably higher than at surface roughness of 0.1 m, which had the values of Cz¼1.0/Cz¼0.1 at the region of 1.08e1.48. Fig. 3 also illustrates that there were significant radon concentration difference between surface roughness of 0.1 m and 1.0 m when the wind speed were lower than 1.0 m s1. At the atmospheric wind speed of 1.0 m s1, 222Rn concentration at surface roughness of 0.1 m dropped by a factor of 2.2 Bq m3 per meter from 50 m to 200 m shown in Fig. 2a, while 222Rn concentration at surface roughness of 1.0 m dropped by a factor of 2.45 Bq m3 per meter from 50 m to 200 m displayed in Fig. 2b. With the increasing distance along downwind direction from shaft outlet exhaust, the values of Cz¼1.0/Cz¼0.1 increased rapidly to the maximum of 1.48 at the distance of 120 m, then dropped remarkably to 1.1 from 120 m
Fig. 2. Contours of radon concentration (Bq.m3) at the height of 1.7 m in the X ¼ 0 plane for surface roughness, of a) 0.1 m, b) 1.0 m.
Fig. 3. Radon concentration comparison in the height of 1.7 m in the X ¼ 0 plane for underlying surface roughness of 0.1 m and 1.0 m.
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
to 200 m. The effects of resistance of surface roughness at high atmospheric wind speeds were not evident as at the low atmospheric wind speeds, and this meant surface roughness and atmospheric wind speeds had a certain negative correlation to frictional resistance. 3.2. Radiological effective dose rate calculation and evaluation As expected, the maximum effective dose rate was found around the shaft outlet, as shown in Fig. 4. At the distance of 50 m downwind from uranium mine ventilation shaft outlet, the 222Rn effective dose rate was very high with the value of 3.0 103 mSv h1. Such high doses which correspond to annual value of 21 mSv at 7000 h y1 exposure time may cause a significant risk to the residents from the health hazard point of view. The value was much higher than the occupational upper limit of 10 mSv y1 recommended by ICRP (ICRP, 2010). The maximum effective dose rates dropped dramatically with the distance changing from 50 m to 200 m, by more than 80%, then reduced slowly further away. The consequence could be explained by the fact that the 222Rn accumulation was caused by the effects of vortex flow at the valley of 120 m along downwind slope surface and then the 222Rn plume would disperse quickly at the location far from the valley. No apparent difference between maximum and minimum value could be found at the distance exceeding 300 m downwind from the uranium mine ventilation shaft. Using the radon radiological effective dose rate calculation methods, 222Rn concentration and effective dose caused by radon at the distance of 250 m, the location of the farm house, were calculated and shown in Table 3. Considering that the farmhouse was located atop uranium-bearing mountains, it was anticipated that indoor radon concentrations would be elevated through ventilation or air permeability, although it was difficult to predict. As a result, the occupants near the active site would receive higher radon dose not only because of exposure to the radon coming directly from the ventilation outlet exhausts, but also exposure to the radon concentrations originated from the rock and soil surrounding the house. A conservative value of 20 h every day of exposure time for the farmer was thus chosen in this study. This represented a total time of 7300 h y1, exceeding the 7000 h y1 indoors or 2000 h y1 at work recommended by ICRP (ICRP, 1993). According to the effective dose calculation equation, the calculated average effective dose rate for a resident at the distance of 250 m was 0.78 mSv y1.
Fig. 4. Maximum and Minimum of public personal effective dose rate per hour E/T (mSv h1) at the height of 1.7 m in the X ¼ 0 plane.
61
Table 3 The average value of 222Rn concentrations and the effective dose rate at Y ¼ 250 m, H ¼ 1.7 m in the X ¼ 0 plane. Atmospheric wind speed (m s1)
Probability
0.5 1.0 2.0 4.0 The total
8% 17% 61% 9%
[CRn] (Bq m3)
E/T (mSv h1)
Z ¼ 0.1
Z ¼ 1.0
Z ¼ 0.1
Z ¼ 1.0
51 30 15 12 19.41
55 32 15.5 12.6 20.43
2.75E-04 1.62E-04 8.1E-05 6.48E-05 1.05E-04
2.97E-04 1.79E-04 8.37E-05 6.8E-05 1.1E-04
The maximum effective dose rate calculated for a resident assumed to be outside at the location of the farm house was 2.2 mSv y1 and it occurred in the conditions of 0.5 m s1 wind speed and 1.0 m surface roughness. The average and maximum effective values calculated in the paper were below the annual effective dose limit of 3e10 mSv recommended by ICRP (ICRP, 1993a, 2010b) for public exposure. In reality, the actual effective dose rate should be lower than the calculated maximum value. The real wind direction towards the farmhouse was not the prevailing wind direction, and the atmospheric wind speed probability at 0.5 m s1 was only 8%. However, in respect of safety, the farmer should show caution and reduce working hours in carrying out occupational activities within the region of 250 m surrounding shaft outlet exhausts. 3.3. Uncertainty analysis of radiological effective dose rate Given the main impact caused by ventilation exhausts to nearby region, the dose assessments for the farmers assumed to be only outside at the location of the farm house. In fact, the different occupancies time outdoors and indoors would cause the exposure time difference between outdoors and indoors such as 10 h outdoors exposure and 12 h indoors exposure according to local regular work hours. Considering ventilation and air permeability, radon concentration indoors were presumed to be higher than outdoors. A conservative value of 20 h every day of exposure time for the farmer was assumed to compensate the exposure indoors in the study. A factor for estimating the ingress of radon activity concentration in air outdoors to indoors needs to be considered on the condition of measurements of outdoor and indoor radon levels in the future work. An equilibrium factor of 0.6 applicable to outdoors was chosen and only outdoor exposure was computed in the paper. Considering that the farmers conducting activities both in outdoors and indoors, so indoor radon exposure needs to be considered at the same time. According to UNSCEAR report (UNSCEAR, 2000), the equilibrium factor applicable to outdoors was ranging from 0.2 to 1.0 while the equilibrium factor applicable to indoors was ranging from 0.1 to 0.9. So there would be different equilibrium factors coupled for indoor and outdoor radon effective rate calculation. Furthermore, the dose conversion factor of 9.0 106 mSv per Bq m3 h was used to calculate the effective dose received by adults per unit 222Rn activity per unit of air volume. But this value may range from 6.0 106 to 15.0 106 mSv per Bq m3 h according to the UNSCEAR report (UNSCEAR, 2000). This would be another uncertainty to calculate radon effective dose rate. Radon diurnal variations in four seasons have been proven by hourly measurements on radon concentrations and the meteorological parameters (Pitari et al., 2013). Atmospheric stability in night time coupled to the temperature inversion at the surface of mountain would discourage vertical air movement and may produce large increases of radon concentrations surrounding to the exhaust outlets, thus would create higher radon concentrations downwind in the valley during morning hours. In addition, atmospheric stability would change in seasons and diurnal cycles under
62
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
mountainous terrains. So radon concentration levels computed in this paper would show diurnal and seasonal variations accordingly. 4. Conclusions 222 Rn dispersion and effective dose analysis in the vicinity area of one uranium mine ventilation shaft were conducted using threedimensional CFD simulations. The investigations of the influence of wind profiles and surface roughness on 222Rn dispersion in the atmosphere revealed that for the 3.0 m s1 shaft exhaust exit velocity and 2.0 m shaft height, the effect of wind speed over 2.0 m s1 on the dispersion process was more pronounced than the effect of underlying surface roughness. Topographic conditions were also found to play a significant role in the dispersion process at low atmospheric wind speeds. The weighting 222Rn concentrations derived from their historical frequency at the distance of 250 m for different wind speeds, and the maximum effective dose value of 2.2 mSv y1, below the low dose limit region of 3e 10 mSv y1 recommended by ICRP (ICRP, 1993a, 2010b), was computed. Due to the radiological soil upon which the farmhouse located, indoor radon may be significant and independent of the outdoor radon concentrations. As the actual radiation dose may vary seasonally, with both indoor radon and the farmer’s outdoor occupational activities, additional work needs to be done in order to refine any dose assessments. Measurements of the levels of radon outdoors and indoors should be carried out in air from both activity concentrations in the air filtering into the building and from radon emanating from the ground below the house. In particular, field measurements indoors and outdoors during different seasons could provide data upon which informed actions aimed at dose reduction could be taken. For more precise and accurate simulations, a larger physical model, more detailed wind profiles, precise radio-aerosol characteristics, better knowledge of the dispersion coefficient of radon, improved characterization of surface roughness, and atmospheric stabilities should be further investigated.
Acknowledgments The authors acknowledge with thanks the financial support provided by National Natural Science Foundation of China (Grant No. 11105068), Postdoctoral Science Foundation of China (Grant No. 2013M542140) and Postdoctoral Science Foundation of Central South University (Grant No.126640). We also deeply thank the anonymous reviewers for their thoughtful suggestions that resulted in a much improved revision of this paper. References Ana Pilar, G.S., Jeroen, V.B., Patrick, R., Domenico, O., 2002. Numerical and experimental modelling of pollutant dispersion in a street canyon. J. Wind Eng. Ind. Aerod. 90, 321e339.
Ansys Fluent 13.0.0, 2010. Ansys Fluent Inc. Basit, A., Espinosa, F., Avila, R., Raza, S., Irfan, N., 2008. Simulation of atmospheric dispersion of radionuclides using an EulerianeLagrangian modelling system. J. Radiol. Prot. 28, 539e561. Brian, E.L., Dudley, B.S., 1972. Lectures in Mathematical Models of Turbulence. Academic Press, London, England. Brian, E.L., Dudley, B.S., 1974. The numerical computation of turbulent flows. Comput. Method. Appl. M 3, 269e289. Bruce, R.W., Werner, S., 1990. Wind-tunnel studies of variable stack heights for a low-profile building. J. Wind Eng. Ind. Aerod. 36, 675e687. Chen, Y.S., Kim, S.W., 1987. Computation of Turbulent Flows Using an Extended k-e Turbulence Closure Model. NASA. CR-179204. de Sampaio, Paulo A.B., Junior, Miltion A.G., Lapa, Celso M.F., 2008. A CFD approach to the atmospheric dispersion of radionuclides in the vicinity of NPPs. Nucl. Eng. Des. 238, 250e273. Dong, X., Hanqing, W., Kimberlee, J.K., 2012. Modeling and experimental validation of the dispersion of 222Rn released from a uranium mine ventilation shaft. Atmos. Environ. 60, 453e459. Evans, R.D., Harley, J.H., Jacobi, W., MacLean, A.S., Mills, W.A., Stewart, C.G., 1981. Estimation of risk from environmental exposure to radon-222 and its decay products. Nature 290, 98e100. Hiroaki, T., Masamichi, C., 2008. Development of an atmospheric dispersion model for accidental discharge of radionuclides with the function of simultaneous prediction for multiple domains and its evaluation by application to the Chernobyl nuclear accident. J. Nucl. Sci. Technol. 45, 920e931. Hong, H., Yoshiaki, A., Mitsuru, A., Masamitsu, T., 2000. A two-dimensional air quality model in an urban street canyon: evaluation and sensitivity analysis. Atmos. Environ. 34, 689e698. Hong, H., Ryozo, O., Shinsuke, K., 2005. Urban thermal environment measurements and numerical simulation for an actual complex urban area covering a large district heating and cooling system in summer. Atmos. Environ. 39, 6362e6375. ICRP (International Commission on Radiological Protection), 1993. Protection Against Radon-222 at Home and at Work 23. Ann. ICRP. ICRP Publication 65. ICRP (International Commission on Radiological Protection), 2010. Lung Cancer Risk from Radon and Progeny and Statement on Radon 40. Ann. ICRP. ICRP Publication 115. Jie, H., Zhiguo, Q., Chihang, Z., Xiangtai, B., 2009. Simulations of pollutant dispersion at toll plazas using three-dimensional CFD models. Trans. Res. D e Tr. E. 14, 557e566. Leelössy, Á., Mészáros, R., Lagzi, I., 2011. Short and long term dispersion patterns of radionuclides in the atmosphere around the Fukushima Nuclear Power Plant. J. Environ. Radioact. 102, 1117e1121. Oliver, G.S., 1953. Micrometeorology. McGraw-Hill, New York. Patankar, S.V., 1980. Numerical Heat Transfer and Fluid Flow. McGraw-Hill, New York. Pitari, Giovanni, Coppari, Eleonora, De Luca, Natalia, Di Carlo, Piero, 2013. Observations and box model analysis of radon-222 in the atmospheric surface layer at L’Aquila, Italy: March 2009 case study. Environ Earth Sci., 1e7. http://dx.doi.org/ 10.1007/s12665-013-2635-1. Qiuju, G., Takao, I., Katsumi, O., 1995. Measurements of thoron concentration by passive cup method and its application to dose assessment. J. Nucl. Sci. Technol. 32, 794e803. Sharma, N., Chaudhry, K.K., Chalapati Rao, C.V., 2005. Air pollution dispersion studies through environmental wind tunnel (EWT) investigations: a review. J. Sci. Ind. Res. 64, 549e559. Srinivas, C.V., Venkatesan, R., 2005. A simulation study of dispersion of air borne radionuclides from a nuclear power plant under a hypothetical accidental scenario at a tropical coastal site. Atmos. Environ. 39, 1497e1511. UNSCEAR, 2000. United Nations Scientific Committee on the Effects of Atomic Radiation, Exposures from Natural Radiation Sources. UNSCEAR 2000 Report to the General Assembly, Annex B. United Nations, New York. UNSCEAR, 2009. UNSCEAR 2006 Report, Annex E. Sources to Effects Assessment for Radon in Homes and Workplaces. United Nations, New York. Victor, Y., Steven, A.O., 1986. Renormalization group analysis of turbulence. J. Sci. Comput. 1, 3e51. Xiaomin, X., Huang, Z., Jiasong, W., 2005. Impact of building configuration on air quality in street canyon. Atmos. Environ. 39, 4519e4530.
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Radon dispersion modeling and dose assessment for uranium mine ventilation shaft exhausts under neutral atmospheric stability Dong Xie a, b, *, Hanqing Wang c, Kimberlee J. Kearfott d, Zehua Liu b, Shunquan Mo b a
School of Energy Science and Engineering, Central South University, Changsha 410083, China School of Urban Construction, University of South China, Hengyang 421001, China c School of Civil Engineering, University of Hunan Technology, Zhuzhou 412007, China d Department of Nuclear Engineering and Radiological Sciences, University of Michigan, Ann Arbor, MI 48109-2104, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 August 2013 Received in revised form 3 December 2013 Accepted 4 December 2013 Available online 28 December 2013
In the present study, the roles of atmospheric wind profiles in the neutral atmosphere and surface roughness parameters in a complex terrain were examined to determine their impacts on radon (222Rn) dispersion from an actual uranium mine ventilation shaft. Simulations were completed on 222Rn dispersion extending from the shaft to a vulnerable distance, near the location of an occupied farmhouse. The eight dispersion scenarios for the ventilation shaft source included four downwind velocities (0.5, 1.0, 2.0 and 4.0 m s1) and two underlying surface roughness characteristics (0.1 m and 1.0 m). 222Rn distributions and elevated pollution regions were identified. Effective dose estimation methods involving a historical weighting of wind speeds in the direction of interest coupled to the complex dispersion model were proposed. Using this approach, the radiation effects on the residents assumed to be outside at the location of the farm house 250 m downwind from the ventilation shaft outlet were computed. The maximum effective dose rate calculated for the residents at the outside of the farm house was 2.2 mSv y1, which is less than the low limit action level of 3e10 mSv y1 recommended by the International Commission on Radiological Protection (ICRP) occupational exposure action level for radon. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: 222 Rn release Atmospheric dispersion Numerical modeling Radiation dose evaluation
1. Introduction Radon (222Rn) emitted from uranium mine ventilation shaft exhausts could constitute a major source of environment contamination and consequently a potential health hazard to the nearby population. Due to their alpha-emitting short lived progeny 218Po and 214Po, 222Rn have long been recognized as main causative agent for lung cancer when presented in high radon inhalation, such as those encountered in uranium mining areas (Evans et al., 1981). The geographical features of the dispersion region and the meteorological conditions are important for evaluating the dispersion of 222 Rn from uranium mine shaft exhausts. Essential parameters to be considered include the 222Rn emission concentration as it leaves the shaft, gas emission velocity, shaft location and height, trees and topography, wind speed and direction, atmospheric stability (Bruce and Werner, 1990) and precipitation.
* Corresponding author. School of Urban Construction, University of South China, Hengyang 421001, China. Tel.: þ86 734 828 2512; fax: þ86 734 828 2312. E-mail address: [email protected] (D. Xie). 0265-931X/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvrad.2013.12.003
Full-scale field measurements, wind tunnel experiments, and computational fluid dynamics (CFD) models have been used to study pollutant dispersion for complex underlying surface conditions. For some field measurement situations, it is hard to simultaneously control operative and intertwined parameters effects (Ana Pilar et al., 2002) such as atmospheric conditions (wind speed, wind direction), topography and geography (underlying surface roughness, mountain height and shaft height, width, shape). The need to utilize hazardous radioactive sources and also the difficulties in creating appropriate boundary conditions similarity may limit the efficiency of wind tunnel experiments (Sharma et al., 2005). However, CFD works well for this situation. CFD has been proved to be a very powerful and efficient tool for the studies of radionuclides dispersion with the factors considered individually or in combination with the wind field effect (de Sampaio et al., 2008). Several previous studies have involved simulations of the atmospheric dispersion of nuclear power plant (NPP) emissions (Srinivas and Venkatesan, 2005; Basit et al., 2008), including from the Chernobyl (Hiroaki and Masamichi, 2008) and Fukushima accidents (Leelössy et al., 2011). 222Rn concentration dispersion and the effective dose evaluation obtained in this study differ from the NPP work not only due to the different radionuclide, but because the
58
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
release was continuously from a low height and the topographical and geographical features and surrounding the release are very different, namely in mountainous rural terrain. Little information was available in the literature about the numerical analysis of 222Rn dispersion and dose evaluation of 222Rn from uranium mine ventilation shafts. Some CFD modeling of the situation has been finished, but it was limited to a single location. This work did, however, included field measurements that provided solid validation for CFD modeling of this problem (Dong et al., 2012). In current study, 222Rn dispersion under neutral atmospheric stability conditions was analyzed using three-dimensional CFD simulation for a specific area above a uranium mine in one province of China. This particular location included an occupied home and farmed area around it. CFD modeling was accomplished using Fluent (Ansys Fluent, 2010), a commercially available and widely used tool incorporating several turbulence models (ANSYS Fluent 13.0.0, ANSYS Inc., Southpointe, 275 Technology Drive, Canonsburg, PA 15317 USA). 222Rn distributions and elevated pollution regions in the surrounding area of uranium mine ventilation shaft were calculated. For this case, public effective dose rate calculation methods were developed and radiation dose rate to the farmers living and working in the vicinity of a house 250 m downwind from the shaft was evaluated. 2. Models and computational methods 2.1. Geometric model The geometric model considered in this study was based upon an actual ventilation shaft located in uranium-bearing mountains in China. Discrete data points for the model were extracted from an elevation contour map and processed with specialized Fluent preprocessing software, Gambit (Ansys Fluent, 2010). The overall computational dimensions of mountain were 400.0 m (L) 400.0 m (W) 150.0 m (H) in the X, Y and Z directions, respectively. The ventilation shaft outlet was modeled to match its actual dimensions of 2.7 m 2.7 m, extending 2.0 m above the ground’s surface. In the present study, Temperature of the exhaust is set to be the same as the atmosphere because of neutral atmospheric stability and large amount of exhaust air rate. 222Rn was emitted from uranium mine ventilation shaft at a certain flow rate of 3.0 m s1, then mixed with air and dispersed in three directions into atmosphere as shown in Fig. 1a. 2.2. Mathematical models Considering that the atmospheric temperature could be regarded as homogeneous with heights in the computation region under the neutral atmospheric stability, CFD simulation was based on the governing equations of continuity, momentum, pollutant transport, turbulent energy and turbulent dissipation rate, as shown in Table 1. Xiaomin et al. (2005) investigated the influence of complex geometry on pollutant dispersion comparing different keε models with wind tunnel measured data for optimization of turbulence models. Their comparison results show that the standard keε model provides the best simulation results (Brian and Dudley, 1974), while Renormalization Group （RNG） keε turbulence models based on RNG theory (Victor and Steven, 1986) and modified keε turbulence models (Chen and Kim, 1987) over-predict the pollutant concentrations. Standard keε closure was thus chosen as the turbulence model in this study. The parameters of equations (1)e(6) depicted in Table 1 are defined as: X is the coordinate axis in the direction i (i ¼ 0, 1, 2), ui corresponds to the mean velocity in i direction, r is the air density, t is time, P is pressure, mt is the turbulent viscosity, m is the molecular viscosity, and g is the
Fig. 1. Fig. 1a. Map of surface above uranium mine showing actually topographic details, including the location of the shaft and the house (250 m downwind), along with the computational slice X ¼ 0 and the southerly wind direction perpendicular to the Y ¼ 0 plane. Fig. 1b. Unstructured computational grid used for finite volume analysis for the CFD model of the actual topographic surface.
gravitational acceleration. K is the turbulence kinetic energy, and ε is the dissipation rate of turbulence kinetic energy. Cm, C1ε, C2ε, sk, and sε are empirical and experimental constants fixed as 0.09, 1.44, 1.92, 1.0, and 1.3, respectively (Brian and Dudley, 1972). C is the 222 Rn average concentration in the air, u is the velocity vector of the 222 Rn, D is the effective diffusion coefficient of 222Rn in air assigned a value of 5 103 m2 s1, which is from the work of Guo (Qiuju et al., 1995), l is the radon decay constant equal to 2.1 106 s1. 2.3. Numerical codes and solution methods CFD modeling was conducted with the code Fluent in this paper. The simulation involved the finite volume discretization of the equations of motion, a geometrical model consisting of an unstructured grid volume made of tetrahedral cells, various matrixinversions routines, and the keε turbulence model (Jie et al., 2009). The coupling between velocity and pressure was accomplished through the SIMPLE algorithm (Patankar, 1980). The central differencing scheme was used in diffusion term and advection term, while an upwind differencing scheme was used in source term. In such a complex terrain area, the unstructured grid system is the most efficient for CFD simulations (Hong et al., 2005), so this approach was selected. The computational domain was built using tetrahedral cells with a finer resolution nearest the ventilation shaft outlet and the mountain ground surface. In the CFD model, a nonuniform tetrahedral grid of approximately 370,000 cells was
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
59
Table 1 Governing equations. Continuity equation vðrui Þ=vXi ¼ 0
(1)
Turbulent momentum equation v rui uj =vXi ¼ vP=vXj þ v ðmt þ mÞ vui =vXj þ vuj =vXi vXi þ rg
(2)
Standard turbulence kinetic energy equation vðrui kÞ=vXi ¼ v½ðm þ mt =sk Þðvk=vXi Þ=vXi þ mt vuj =vXi vui =vXj þ vuj =vXi rε
(3)
Standard dissipation rate of turbulence kinetic energy vðrui εÞ=vXi ¼ v½ðm þ mt =sε Þðvk=vXi Þ=vXi þ C1ε mt ðε=kÞ vuj =vXi vui =vXj þ vuj =vXi rC2ε ε2 =k
mt ¼ Cm r k2 =ε
(4)
(5)
Radon transport equation vðCuÞ=vXi ¼ D v2 C=vXi2 þ Q lC
(6)
created. The triangular mesh of the computational domain of the topographic surface was shown in Fig. 1b. 2.4. Boundary conditions and parameters Inlet boundary condition was used for the atmospheric wind with specified velocity and direction. Simulations were performed for atmospheric wind velocities of 0.5, 1.0, 2.0 and 4.0 m s1 for class D, or neutral, atmosphere stability condition. For the location of the ventilation shaft, neutral atmosphere stability condition occurs 52.1% of the time according to local meteorological conditions. The probabilities of atmospheric wind speeds of 0.5,1.0, 2.0, 4.0 m s1 are 8%, 17%, 61%, 9% and 5%, respectively. The remaining 5% is the wind speed above 4.0 m s1 and its influence is neglected in this study. A user-defined subroutine compiled by Cþþ was developed and used in the analysis for boundary conditions, for example, power law velocity profiles applied at atmospheric wind inlet and turbulence kinetic energy and dissipation rate applied at shaft outlet with the Fluent code. The turbulence kinematic energy k and dissipation rate ε at the shaft outlet were defined in Table 2. The parameters of equations (7)e(9) depicted in Table 2 are defined as: uY is the wind speed at height Z, u0 is the average wind speed at the height of a reference height Z0 (Z0 ¼ 10 m), and the exponent a is a function of both the atmosphere stability in the layer and the underlying surface characteristics. u is the average wind velocity, I is fractional turbulent component, and L is turbulence integration scale. Table 2 Boundary conditions and calculation parameters settings. Wind inlet uY =u0 ¼ ðZ=Z0 Þa ; Z0 ¼ 10m
(7)
z ¼ 0.1 m, a ¼ 0.19 z ¼ 1.0 m, a ¼ 0.25 (Oliver, 1953)
Shaft outlet k ¼ 1:5ðu IÞ2 ε ¼ Cm0:75 k1:5 =l;
l ¼ 4ð0:09kÞ0:5 Z00:25 Z 0:75 =U0
Side and top Wall Surface roughness
(8)
ðHong et al:; 2005Þ Outflow Non-equilibrium wall function z ¼ 0.1 m, 1.0 m
(9)
Another inlet boundary condition was the radon-bearing exhaust jet plume from the shaft outlet. The shaft exhauss velocity was set to its average measured value of 3.0 m s1. 222Rn concentration was measured at the side of the top of the ventilation shaft to be 8000 Bq m3 and the value was used as the source term. For the outlets at the end, the top and the lateral sides of the domain, outflow boundary conditions were assumed. The ground surface and the lateral sides of the shaft outlet were treated as walls with no-slip velocity boundary condition, and a non-equilibrium wall function was applied for near wall treatment (Jie et al., 2009). The wind blowed in the southerly direction towards the farmer’s house occurs in 25% of the time. Based upon local topographical and geomorphologic features across the mountain being modeled, as well as the effects of seasons for spring-summer and autumnwinter, two underlying surface roughness (0.1 m, 1.0 m) were considered in the study. In this study, surface roughness of 1.0 m represented the seasons effects of spring-summer as there were many vegetations like leaves on trees or grass on the ground, which would pose relative higher underlying surface frictional resistance. The surface roughness of 0.1 m represented the less vegetation effects of seasons for autumn-winter. 2.5. Radiological dose calculation methods Using the radon concentration values obtained from CFD calculation, the effective dose rate for the residents assumed to be outside at the location of the farm house was calculated with standardized methods (UNSCEAR, 2000). The corresponding dose conversion factor value, i.e. the effective dose rate received by adults per unit 222Rn activity per unit of air volume, used for the computation, 9.0 106 mSv per Bq m3 h, was remained constant for several years (UNSCEAR, 2000a, 2009b). The effective dose rate to the public from radon at different distances from ventilation shaft outlet was estimated using following equation:
E=T ¼ C F D
(10)
where, E/T was public personal effective dose rate (mSv h1), C was radon concentration (Bq m3), F was an equilibrium factor applicable to outdoors and it was 0.6 found from UNSCEAR report (UNSCEAR, 2000), T was the annual exposing hours obtained from
60
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
the questionnaires finished by the occupants who lived and worked in the environment of interest (h y1), and D was the dose conversion factor (9.0 106 mSv per Bq m3 h). Considering that atmospheric wind speeds probability distribution, the average effective dose rate posed by 222Rn could be depicted as:
ETotal =T ¼
X
fV CV F D
(11)
V
where, ETotal/T was the sum of the public personal effective dose rate per hour (mSv h1), CV was radon concentration under atmospheric wind speeds V of 0.5, 1.0, 2.0, 4.0 m s1 (Bq m3), fV was the fractional probability of a given atmospheric wind speed. 3. Results and discussion 3.1. Radon concentration distribution For the present simulation, the exhaust emissions went downwind and the exhaust emissions would be trapped in the valley which located at the distance of 120 m from shaft outlet exhausts. The effect was caused by that the shaft height was relatively low at a certain shaft exhausts outlet rate and the location of the shaft was located on the hillsides of mountains. Due to the valley accumulation effects, 222Rn concentration with the region of 150 m on the downstream direction were high and the plume cannot transport efficiently, as shown in Fig. 2. When ambient wind speed was 0.5 m s1, the diffusion ability was not significant. 222Rn concentration from shaft outlet exhausts was higher than 500 Bq m3 within the region of 100 m. When atmospheric wind speed increased from 0.5 m s1 to 1.0 m s1, 222 Rn concentration at surface roughness of 0.1 m dropped from 500e378 Bq m3, while 222Rn concentration at surface roughness of 1.0 m dropped from 556e390 Bq m3 at the distance of 50 m. As shown in Fig. 2, ambient wind profiles played the dominant effect in radon dispersion process. Besides wind direction, wind speed was another important factor as higher wind speeds leaded to more effective pollutants dilution (Hong et al., 2000). Fig. 2 also indicates that the effect of surface roughness was negligible when the atmospheric wind speeds exceeded 2.0 m s1. The radioactive pollution along downwind direction on local region over 200 m was within the limits of 100 Bq m3, as the dispersion ability of 222Rn was strong and 222Rn could travel further at higher atmospheric wind speeds. Compared with predicted downstream radon concentrations for an assumed surface roughness of 0.1 m, 222Rn concentrations for simulated surface roughness of 1.0 m were larger, as shown in Fig. 3. An increased number of plants and trees, associated with a high surface roughness, were expected to cause the accumulation of 222 Rn. As shown at about 120 m from the exhaust outlet in the downwind direction, radon concentrations at surface roughness of 1.0 m were remarkably higher than at surface roughness of 0.1 m, which had the values of Cz¼1.0/Cz¼0.1 at the region of 1.08e1.48. Fig. 3 also illustrates that there were significant radon concentration difference between surface roughness of 0.1 m and 1.0 m when the wind speed were lower than 1.0 m s1. At the atmospheric wind speed of 1.0 m s1, 222Rn concentration at surface roughness of 0.1 m dropped by a factor of 2.2 Bq m3 per meter from 50 m to 200 m shown in Fig. 2a, while 222Rn concentration at surface roughness of 1.0 m dropped by a factor of 2.45 Bq m3 per meter from 50 m to 200 m displayed in Fig. 2b. With the increasing distance along downwind direction from shaft outlet exhaust, the values of Cz¼1.0/Cz¼0.1 increased rapidly to the maximum of 1.48 at the distance of 120 m, then dropped remarkably to 1.1 from 120 m
Fig. 2. Contours of radon concentration (Bq.m3) at the height of 1.7 m in the X ¼ 0 plane for surface roughness, of a) 0.1 m, b) 1.0 m.
Fig. 3. Radon concentration comparison in the height of 1.7 m in the X ¼ 0 plane for underlying surface roughness of 0.1 m and 1.0 m.
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
to 200 m. The effects of resistance of surface roughness at high atmospheric wind speeds were not evident as at the low atmospheric wind speeds, and this meant surface roughness and atmospheric wind speeds had a certain negative correlation to frictional resistance. 3.2. Radiological effective dose rate calculation and evaluation As expected, the maximum effective dose rate was found around the shaft outlet, as shown in Fig. 4. At the distance of 50 m downwind from uranium mine ventilation shaft outlet, the 222Rn effective dose rate was very high with the value of 3.0 103 mSv h1. Such high doses which correspond to annual value of 21 mSv at 7000 h y1 exposure time may cause a significant risk to the residents from the health hazard point of view. The value was much higher than the occupational upper limit of 10 mSv y1 recommended by ICRP (ICRP, 2010). The maximum effective dose rates dropped dramatically with the distance changing from 50 m to 200 m, by more than 80%, then reduced slowly further away. The consequence could be explained by the fact that the 222Rn accumulation was caused by the effects of vortex flow at the valley of 120 m along downwind slope surface and then the 222Rn plume would disperse quickly at the location far from the valley. No apparent difference between maximum and minimum value could be found at the distance exceeding 300 m downwind from the uranium mine ventilation shaft. Using the radon radiological effective dose rate calculation methods, 222Rn concentration and effective dose caused by radon at the distance of 250 m, the location of the farm house, were calculated and shown in Table 3. Considering that the farmhouse was located atop uranium-bearing mountains, it was anticipated that indoor radon concentrations would be elevated through ventilation or air permeability, although it was difficult to predict. As a result, the occupants near the active site would receive higher radon dose not only because of exposure to the radon coming directly from the ventilation outlet exhausts, but also exposure to the radon concentrations originated from the rock and soil surrounding the house. A conservative value of 20 h every day of exposure time for the farmer was thus chosen in this study. This represented a total time of 7300 h y1, exceeding the 7000 h y1 indoors or 2000 h y1 at work recommended by ICRP (ICRP, 1993). According to the effective dose calculation equation, the calculated average effective dose rate for a resident at the distance of 250 m was 0.78 mSv y1.
Fig. 4. Maximum and Minimum of public personal effective dose rate per hour E/T (mSv h1) at the height of 1.7 m in the X ¼ 0 plane.
61
Table 3 The average value of 222Rn concentrations and the effective dose rate at Y ¼ 250 m, H ¼ 1.7 m in the X ¼ 0 plane. Atmospheric wind speed (m s1)
Probability
0.5 1.0 2.0 4.0 The total
8% 17% 61% 9%
[CRn] (Bq m3)
E/T (mSv h1)
Z ¼ 0.1
Z ¼ 1.0
Z ¼ 0.1
Z ¼ 1.0
51 30 15 12 19.41
55 32 15.5 12.6 20.43
2.75E-04 1.62E-04 8.1E-05 6.48E-05 1.05E-04
2.97E-04 1.79E-04 8.37E-05 6.8E-05 1.1E-04
The maximum effective dose rate calculated for a resident assumed to be outside at the location of the farm house was 2.2 mSv y1 and it occurred in the conditions of 0.5 m s1 wind speed and 1.0 m surface roughness. The average and maximum effective values calculated in the paper were below the annual effective dose limit of 3e10 mSv recommended by ICRP (ICRP, 1993a, 2010b) for public exposure. In reality, the actual effective dose rate should be lower than the calculated maximum value. The real wind direction towards the farmhouse was not the prevailing wind direction, and the atmospheric wind speed probability at 0.5 m s1 was only 8%. However, in respect of safety, the farmer should show caution and reduce working hours in carrying out occupational activities within the region of 250 m surrounding shaft outlet exhausts. 3.3. Uncertainty analysis of radiological effective dose rate Given the main impact caused by ventilation exhausts to nearby region, the dose assessments for the farmers assumed to be only outside at the location of the farm house. In fact, the different occupancies time outdoors and indoors would cause the exposure time difference between outdoors and indoors such as 10 h outdoors exposure and 12 h indoors exposure according to local regular work hours. Considering ventilation and air permeability, radon concentration indoors were presumed to be higher than outdoors. A conservative value of 20 h every day of exposure time for the farmer was assumed to compensate the exposure indoors in the study. A factor for estimating the ingress of radon activity concentration in air outdoors to indoors needs to be considered on the condition of measurements of outdoor and indoor radon levels in the future work. An equilibrium factor of 0.6 applicable to outdoors was chosen and only outdoor exposure was computed in the paper. Considering that the farmers conducting activities both in outdoors and indoors, so indoor radon exposure needs to be considered at the same time. According to UNSCEAR report (UNSCEAR, 2000), the equilibrium factor applicable to outdoors was ranging from 0.2 to 1.0 while the equilibrium factor applicable to indoors was ranging from 0.1 to 0.9. So there would be different equilibrium factors coupled for indoor and outdoor radon effective rate calculation. Furthermore, the dose conversion factor of 9.0 106 mSv per Bq m3 h was used to calculate the effective dose received by adults per unit 222Rn activity per unit of air volume. But this value may range from 6.0 106 to 15.0 106 mSv per Bq m3 h according to the UNSCEAR report (UNSCEAR, 2000). This would be another uncertainty to calculate radon effective dose rate. Radon diurnal variations in four seasons have been proven by hourly measurements on radon concentrations and the meteorological parameters (Pitari et al., 2013). Atmospheric stability in night time coupled to the temperature inversion at the surface of mountain would discourage vertical air movement and may produce large increases of radon concentrations surrounding to the exhaust outlets, thus would create higher radon concentrations downwind in the valley during morning hours. In addition, atmospheric stability would change in seasons and diurnal cycles under
62
D. Xie et al. / Journal of Environmental Radioactivity 129 (2014) 57e62
mountainous terrains. So radon concentration levels computed in this paper would show diurnal and seasonal variations accordingly. 4. Conclusions 222 Rn dispersion and effective dose analysis in the vicinity area of one uranium mine ventilation shaft were conducted using threedimensional CFD simulations. The investigations of the influence of wind profiles and surface roughness on 222Rn dispersion in the atmosphere revealed that for the 3.0 m s1 shaft exhaust exit velocity and 2.0 m shaft height, the effect of wind speed over 2.0 m s1 on the dispersion process was more pronounced than the effect of underlying surface roughness. Topographic conditions were also found to play a significant role in the dispersion process at low atmospheric wind speeds. The weighting 222Rn concentrations derived from their historical frequency at the distance of 250 m for different wind speeds, and the maximum effective dose value of 2.2 mSv y1, below the low dose limit region of 3e 10 mSv y1 recommended by ICRP (ICRP, 1993a, 2010b), was computed. Due to the radiological soil upon which the farmhouse located, indoor radon may be significant and independent of the outdoor radon concentrations. As the actual radiation dose may vary seasonally, with both indoor radon and the farmer’s outdoor occupational activities, additional work needs to be done in order to refine any dose assessments. Measurements of the levels of radon outdoors and indoors should be carried out in air from both activity concentrations in the air filtering into the building and from radon emanating from the ground below the house. In particular, field measurements indoors and outdoors during different seasons could provide data upon which informed actions aimed at dose reduction could be taken. For more precise and accurate simulations, a larger physical model, more detailed wind profiles, precise radio-aerosol characteristics, better knowledge of the dispersion coefficient of radon, improved characterization of surface roughness, and atmospheric stabilities should be further investigated.
Acknowledgments The authors acknowledge with thanks the financial support provided by National Natural Science Foundation of China (Grant No. 11105068), Postdoctoral Science Foundation of China (Grant No. 2013M542140) and Postdoctoral Science Foundation of Central South University (Grant No.126640). We also deeply thank the anonymous reviewers for their thoughtful suggestions that resulted in a much improved revision of this paper. References Ana Pilar, G.S., Jeroen, V.B., Patrick, R., Domenico, O., 2002. Numerical and experimental modelling of pollutant dispersion in a street canyon. J. Wind Eng. Ind. Aerod. 90, 321e339.
Ansys Fluent 13.0.0, 2010. Ansys Fluent Inc. Basit, A., Espinosa, F., Avila, R., Raza, S., Irfan, N., 2008. Simulation of atmospheric dispersion of radionuclides using an EulerianeLagrangian modelling system. J. Radiol. Prot. 28, 539e561. Brian, E.L., Dudley, B.S., 1972. Lectures in Mathematical Models of Turbulence. Academic Press, London, England. Brian, E.L., Dudley, B.S., 1974. The numerical computation of turbulent flows. Comput. Method. Appl. M 3, 269e289. Bruce, R.W., Werner, S., 1990. Wind-tunnel studies of variable stack heights for a low-profile building. J. Wind Eng. Ind. Aerod. 36, 675e687. Chen, Y.S., Kim, S.W., 1987. Computation of Turbulent Flows Using an Extended k-e Turbulence Closure Model. NASA. CR-179204. de Sampaio, Paulo A.B., Junior, Miltion A.G., Lapa, Celso M.F., 2008. A CFD approach to the atmospheric dispersion of radionuclides in the vicinity of NPPs. Nucl. Eng. Des. 238, 250e273. Dong, X., Hanqing, W., Kimberlee, J.K., 2012. Modeling and experimental validation of the dispersion of 222Rn released from a uranium mine ventilation shaft. Atmos. Environ. 60, 453e459. Evans, R.D., Harley, J.H., Jacobi, W., MacLean, A.S., Mills, W.A., Stewart, C.G., 1981. Estimation of risk from environmental exposure to radon-222 and its decay products. Nature 290, 98e100. Hiroaki, T., Masamichi, C., 2008. Development of an atmospheric dispersion model for accidental discharge of radionuclides with the function of simultaneous prediction for multiple domains and its evaluation by application to the Chernobyl nuclear accident. J. Nucl. Sci. Technol. 45, 920e931. Hong, H., Yoshiaki, A., Mitsuru, A., Masamitsu, T., 2000. A two-dimensional air quality model in an urban street canyon: evaluation and sensitivity analysis. Atmos. Environ. 34, 689e698. Hong, H., Ryozo, O., Shinsuke, K., 2005. Urban thermal environment measurements and numerical simulation for an actual complex urban area covering a large district heating and cooling system in summer. Atmos. Environ. 39, 6362e6375. ICRP (International Commission on Radiological Protection), 1993. Protection Against Radon-222 at Home and at Work 23. Ann. ICRP. ICRP Publication 65. ICRP (International Commission on Radiological Protection), 2010. Lung Cancer Risk from Radon and Progeny and Statement on Radon 40. Ann. ICRP. ICRP Publication 115. Jie, H., Zhiguo, Q., Chihang, Z., Xiangtai, B., 2009. Simulations of pollutant dispersion at toll plazas using three-dimensional CFD models. Trans. Res. D e Tr. E. 14, 557e566. Leelössy, Á., Mészáros, R., Lagzi, I., 2011. Short and long term dispersion patterns of radionuclides in the atmosphere around the Fukushima Nuclear Power Plant. J. Environ. Radioact. 102, 1117e1121. Oliver, G.S., 1953. Micrometeorology. McGraw-Hill, New York. Patankar, S.V., 1980. Numerical Heat Transfer and Fluid Flow. McGraw-Hill, New York. Pitari, Giovanni, Coppari, Eleonora, De Luca, Natalia, Di Carlo, Piero, 2013. Observations and box model analysis of radon-222 in the atmospheric surface layer at L’Aquila, Italy: March 2009 case study. Environ Earth Sci., 1e7. http://dx.doi.org/ 10.1007/s12665-013-2635-1. Qiuju, G., Takao, I., Katsumi, O., 1995. Measurements of thoron concentration by passive cup method and its application to dose assessment. J. Nucl. Sci. Technol. 32, 794e803. Sharma, N., Chaudhry, K.K., Chalapati Rao, C.V., 2005. Air pollution dispersion studies through environmental wind tunnel (EWT) investigations: a review. J. Sci. Ind. Res. 64, 549e559. Srinivas, C.V., Venkatesan, R., 2005. A simulation study of dispersion of air borne radionuclides from a nuclear power plant under a hypothetical accidental scenario at a tropical coastal site. Atmos. Environ. 39, 1497e1511. UNSCEAR, 2000. United Nations Scientific Committee on the Effects of Atomic Radiation, Exposures from Natural Radiation Sources. UNSCEAR 2000 Report to the General Assembly, Annex B. United Nations, New York. UNSCEAR, 2009. UNSCEAR 2006 Report, Annex E. Sources to Effects Assessment for Radon in Homes and Workplaces. United Nations, New York. Victor, Y., Steven, A.O., 1986. Renormalization group analysis of turbulence. J. Sci. Comput. 1, 3e51. Xiaomin, X., Huang, Z., Jiasong, W., 2005. Impact of building configuration on air quality in street canyon. Atmos. Environ. 39, 4519e4530.
