### Introduction

*ΔT*) method and the standard deviation of horizontal wind direction fluctuations (

*σ*

*) method are the general methods for determining atmospheric stability [3]. The Korean Nuclear Safety and Security Commission Notification No. 2014-25 concerning the assessment standards of the meteorological conditions of nuclear facilities follows and endorses the NRC guidelines for classifying atmospheric stability [4]. Methods of classifying and evaluating atmospheric stability differ among countries; nuclear facilities in Korea principally use the*

_{θ}*ΔT*method to estimate atmospheric stability and the

*σ*

*method as an ancillary method. In cases where the meteorological data required for assessing*

_{θ}*ΔT*cannot be obtained due to the limited height of meteorological tower, the

*σ*

*method must be used as an alternative.*

_{θ}*ΔT*or the

*σ*

*method.*

_{θ}### Materials and Methods

### 1. Analysis of the meteorological data

*ΔT*), standard deviation of horizontal wind direction fluctuation (

*σ*

*), solar radiation, and cloud cover [8].*

_{θ}*ΔT*or

*σ*

*(Table 1). The Pasquill system is a fairly reliable reflection of synoptic atmospheric stability under steady-state conditions, and because the routine collection of meteorological elements is useful for other purposes, this classification system is widely used. Our assessment used 7 subclasses of stability, with category A representing extremely unstable conditions and category G representing extremely stable conditions. The*

_{θ}*ΔT*method of classification of atmospheric stability uses the difference in temperature between two atmospheric levels, and takes into account vertical mixing in the atmosphere and the vertical dispersal induced by wind velocity. The meteorological measurements involved in the

*ΔT*method are useful and straightforward for estimating atmospheric stability, but require an observatory tower of a certain height. Cramer proposed the

*σ*

*method for estimating atmospheric stability, which is based on the hypothesis that the dispersion of atmospheric pollutants induced by strong wind conditions at a given altitude is similar to the dispersion caused during conditions of atmospheric instability [9]. Similarly to the*

_{θ}*ΔT*measurements, because the

*σ*

*data have been shown to be useful in other ways and the*

_{θ}*σ*

*has indeed been shown to be closely correlated to atmospheric stability, its use is widespread.*

_{θ}*ΔT*and

*σ*

*to determine Pasquill stability classes at the Daedeok site and the Wolsong site. To calculate*

_{θ}*ΔT*, the temperatures were measured at 10 and 67 meters above ground level at the Daedeok site and at 10 and 58 meters above ground level at the Wolsong site, while

*σ*

*was calculated from measurements made 10 meters above ground level. The resulting parameters were used to estimate and classify atmospheric stability.*

_{θ}### 2. Annual averaged atmospheric dispersion and ground deposition

*Q*) and the ground deposition factor (

*D*/

*Q*), respectively. In accordance to the NRC Regulatory Guide 1.111, a straight-line Gaussian plume model using a statistical data that is compiled into the annual joint frequency distributions according to atmospheric stability class can be used to simulate dispersion. With regard to ground-level release, the higher of the atmospheric dispersion factors derived from equations (1) and (2) is used.

##### (1)

$$\frac{\chi}{Q}(x,k)=\frac{2.032}{x}RF(x,k)\sum _{ij}D{P}_{ij}(x,k)D{C}_{i}(x){f}_{ij}(x,k){[{u}_{i}{({\sigma}_{zj}^{2}(x)+C{D}_{Z}^{2}/\pi )}^{1/2}]}^{-1}$$##### (2)

$$\frac{\chi}{Q}(x,k)=\frac{2.032}{x}RF(x,k)\sum _{ij}D{P}_{ij}(x,k)D{C}_{i}(x){f}_{ij}(k){[\sqrt{3}{u}_{i}{\sigma}_{zj}(x)]}^{-1}$$χ/

*Q*: annual averaged atmospheric dispersion factor (sec·m^{−3})*x*: downward wind distance (m)*i*: index of wind speed*j*: index of atmospheric stability classes*k*: index of wind direction*u*: wind velocity along the plume centerline (m·sec^{−1})*σ*: vertical diffusion coefficient (m)_{zj}*f*: probability of meteorological occurrence*RF*: correction factor for air recirculation and stagnation*DP*: correction factor for plume depletion*DC*: correction factor for radioactive decay*C*: constant for building wake effect (=0.5)*D*: structure height [m]_{z}$2.032={\scriptstyle \frac{2n}{{(2\pi )}^{3/2}}}$ (

*n*: number of wind-direction sectors)

### Results and Discussion

*ΔT*or

*σ*

*, we performed a comparative analysis of the meteorological characteristics derived from these two methods of classifying stability. Assuming a ground level release, we used meteorological data obtained 10 meters above ground level. For the assessment of*

_{θ}*ΔT*, measurements were taken at 10 and 67 meters above ground level at Daedeok and at 10 and 58 meters above ground level at Wolsong. Other meteorological characteristics, such as the building wake effect and the recirculation and stagnation of building turbulence that are specifications of each nuclear facility were not considered. Rather, in this study, we chose only to investigate the effects of meteorological characteristics on atmospheric stability. In accordance with the NRC Regulatory Guide 1.23, we subdivided wind speed into 11 categories.

*ΔT*or

*σ*

*) at the Daedeok site and at the Wolsong site. For the Daedeok site, the distribution of atmospheric stability results based on*

_{θ}*ΔT*was generally homogeneous, showing a slight peak at category E (slightly stable atmospheric conditions). In contrast, the distribution of atmospheric stability results based on

*σ*

*showed a highly skewed distribution; the occurrence of category A (extremely unstable) atmospheric conditions had a probability over 90%. For the Wolsong site,*

_{θ}*ΔT*-based atmospheric stability was most strongly associated with category E (slightly stable) atmospheric conditions and

*σ*

*-based atmospheric stability was most strongly associated with category A (extremely unstable) atmospheric conditions. However, these associations were not as strong as those observed for the Daedeok site.*

_{θ}*ΔT*was 1.46×10

^{−4}sec·m

^{−3}in the east direction, and that based on

*σ*

*was 2.25×10*

_{θ}^{1}sec·m

^{−3}in the east-northeast direction. Likewise, Figures 6 and 7 illustrate the atmospheric dispersion factors according to the classification method at various calculation distances of the Wolsong site. The maximum atmospheric dispersion factor based on

*ΔT*was 6.62×10

^{−5}sec·m

^{−3}in the east-southeast direction and the maximum atmospheric dispersion factor based on

*σ*

*was 1.50×10*

_{θ}^{−5}sec·m

^{−3}in the east-southeast direction. The maximum values of the ground deposition factor, regardless of the classification method used, were found in the east-northeast direction at the Daedeok site (Figures 8, 9) and in the east-southeast direction at the Wolsong site (Figures 10, 11). The maximum values of the ground deposition factor were as follows: at the Daedeok site, 6.24×10

^{−8}m

^{−3}(ΔT method) and 6.28×10

^{−8}m

^{−3}(σ

_{θ}method); and at the Wolsong site, 7.92×10

^{−8}m

^{−3}(ΔT method) and 7.76×10

^{−8}m

^{−3}(σ

_{θ}method). Our results suggest that atmospheric stability based on

*ΔT*is the more conservative in calculating the annual averaged atmospheric dispersion factor and the ground deposition factor, with the exception of the ground deposition factor at the Daedeok site.

*ΔT*was higher than that based on

*σ*

*. The difference between the two atmospheric dispersion factors by different classification systems appears to be correlated to the distance from the nuclear power station: the greater the distance from the nuclear facilities, the larger the difference between the factors. The atmospheric dispersion factor based on*

_{θ}*ΔT*was 6.5 to 14.5 times higher than the atmospheric dispersion factor based on

*σ*

*at the Daedeok site and 4.4 to 6.8 times higher at the Wolsong site.*

_{θ}*ΔT*to ground deposition factor based on

*σ*

*) was not influenced by distance. The corresponding ratios were almost same for the Daedeok and Wolsong sites. The reason that the impact of the classification method on the atmospheric dispersion factor was found to be sensitive to the distance from the nuclear site, whereas no distance dependency was observed for the ground deposition factor, may be because the ground deposition factor is evaluated through the relative deposition rate, which takes distance into account, and because ground-level release is calculated independently of atmospheric stability.*

_{θ}### Conclusion

*ΔT*or

*σ*

*, we analyzed the atmospheric dispersion factor and the ground deposition factor at the Daedeok site and the Wolsong site. For both sites, we found that the atmospheric dispersion factor derived from the atmospheric stability class obtained using*

_{θ}*ΔT*was higher than that derived from the atmospheric stability class obtained using

*σ*

*. For the ground deposition factor, since the ground-level release, which is estimated through empirical graphs presented in the NRC Regulatory Guide 1.111, is assessed independently of atmospheric stability, the results of the two methods were the same. Since our study was an investigation for selecting sites in a specific year, it is difficult to say that the findings of our study have strong implications for different cases. However, our findings suggest that in general,*

_{θ}*ΔT*is the more conservative method for offsite dose calculations for nuclear facilities. If at the time of preliminary study for the determination of site distance, no local meteorological tower is tall enough to measure

*ΔT*,

*σ*

*may have to be measured. In this case, the margin for site distance should be considered to meet the national standards to protect radiation-induced health detriment to the public.*

_{θ}