There are three pathways by which a person can be exposed to radiation from contaminated ground. These are the pathway of radiation exposure caused by the contaminated soil, the pathway of radiation exposure caused by possible ingestion or skin contact with scattered soil particles and the pathway of internal radiation exposure caused by eating farm produce that has adsorbed radioactive substances from contaminated ground (Tsukada et al., 1998), as shown in Figure 1.

Figure 1 Pathways through which a person is exposed to radiation from contaminated ground

According to a study that compared radiation exposure resulting from contaminated soils with the internal radiation exposure caused by ingestion or skin contact with scattered soil particles, the latter represents 95% of all radiation exposure (Ministry of Education, Science and Culture, 2011b). The Ministry of Health, Labour and Welfare has been performing dosimetry of the internal radiation caused by consumption of farm produce that has absorbed radioactive substances from contaminated soil. Food monitoring was performed from March to August 2011. According to the results, the dosage in the farm produce was 0·041–0·135 mSv over 6 months (Ministry of Health, Labour and Welfare, 2011), whereas a person would receive a dosage of approximately 1·2 mSv by consuming dairy products over 6 months (UNSCEAR, 1988). Therefore, the soil can be evaluated as safe in terms of human health.

Reduction methods for the pathways of radiation exposure resulting from contaminated soil can be classified into four measures (Tsukada et al., 1998), as shown in Figure 2. These are system management, original position management, original position processing and removal of the contaminants. Figure 2 shows a schematic diagram of the classification of the four measures. Furthermore, each measure is outlined and discussed in turn below.

Figure 2 The classification of reduction methods for contaminated grounds

System management is a method of reducing dosage by forbidding people to enter the contaminated area. Prohibiting entrance to a contaminated area limits the measures required by designating a remediated evacuation zone and caution area. The technical degree of difficulty is low; nevertheless, land use is extremely limited. Since the residents of a contaminated area must be moved, the mental and financial burdens of each refugee, as well as the social cost and limitations on land use, can become significant.

Original position management is a method of reducing the dosage that occurs as a result of falling air radiation contamination by covering the ground surface and reverse tillage. The technical degree of difficulty of this method is also low, and the cost of implementation is inexpensive. However, long-term management is necessary, and land use must be limited in a controlled manner.

Original position processing is a method of decreasing the concentration of radioactive substances at the field site, which does not involve removing the contaminated ground. Such methods include soil disturbance, removal by water using paddy fields and phytoremediation.

The removal of contaminants is a method of decreasing the concentration of radioactive substances at the field site by removing contaminated ground. This method consists of three phases: removing the contaminated ground, volume reduction and storage. After removing contaminated ground, there is a configuration for storing the contaminated ground without reducing the volume and a configuration for storing the concentrate that is composed of the radioactive substances removed from the contaminated ground.

These measures differ in terms of their direct costs, period required for implementation, technical completeness of the solution and indirect costs.

Concentration decrement by radioactive decay depends on the radioactive half-life. Figure 4 shows the time courses for the survival rates of iodine-131, caesium-131 and caesium-137. The radioactive half-life of iodine-131 is almost 8 d. Therefore, its level of contamination becomes negligible relatively quickly. Caesium-131, whose radioactive half-life is approximately 2 years, disappears after approximately 15 years. However, the radioactive half-life of caesium-137 is almost 30 years, requiring long-term scrutiny.

Figure 4 Circumstances of decay by radioactive half-life

The existing form of radioactive caesium and the soil properties are important factors to consider. Radioactive caesium in soil exists as a monovalent cation, Cs⁺. There are clay minerals and organic substances that have a significant amount of negative electric charge on their surface. Therefore, radioactive caesium is adsorbed on them in a similar manner to other heavy metals and positive ions (Nguyen et al., 2017).

Adsorption of radioactive caesium with frayed edge site

A 2:1 type layer silicate is one of the mineral constituents of clay particles. The silicate holds negative charges on a silicon tetrahedral sheet. Mica holds a larger number of negative charges than potassium ion (K⁺) and occupies a six-membered ring. Consequently, other negative charges can enter the layer as the layer interval already has been strongly closed.

A distal portion of the layer swells in the state that maintained the seat structure by aeration. As a result, potassium ion is released from a layer interval of a circumferential section and a hydrated positive ion is held in a layer interval. The wedge-shaped border section of the swelling and non-swelling layers is called the frayed edge site. In a frayed edge site, a hydrated positive ion is released by spatial limitation. Then, caesium ions, whose hydration energy is the lowest and whose shape is fit for a six-membered ring, is adsorbed strongly to the frayed edge site, as can be seen in Figure 5.

Figure 5 Edge structure of weathered mica

Absorption by plants

There are two possibilities for radioactive substances released into the atmosphere to be transferred to plants. One pathway is directly from the atmosphere, which is called the directly absorbed course (Mohammad, 2015). The other pathway is when radioactive substances that have fallen on the ground are absorbed into the root of the plant, which is called the root absorption course, as shown in Figure 6.

Figure 6 The courses of the transfer of radioactive substances into a plant

Following the Fukushima Daiichi Nuclear Power Plant accident, the directly absorbed course was a significant influence when the quantity of radioactive substances was large. Afterwards, the root absorption course became more influential. Furthermore, since pollution of the plants by the directly composed course was dependent on the quantity of radioactive substances, the period was relatively short. In contrast, most radioactive substances remained in the soil for a long period of time because of their strong absorption characteristics. Therefore, radioactive pollution of the plants by the root absorption course is expected to last several decades owing to the long radioactive half-life.

The quantity of absorption by the root absorption course is dependent on the concentration of the radioactive substances and the transition coefficient. The transition coefficient is a parameter measuring the tendency of radioactive substances to transfer from soil to plants and can be written as

$$\text{transition\hspace{0.17em}coefficient}=\frac{\text{concentration\hspace{0.17em}of\hspace{0.17em}the\hspace{0.17em}radioactive\hspace{0.17em}substance\hspace{0.17em}in\hspace{0.17em}edible\hspace{0.17em}part\hspace{0.17em}of\hspace{0.17em}plant}}{\text{concentration\hspace{0.17em}of\hspace{0.17em}the\hspace{0.17em}radioactive\hspace{0.17em}substance \hspace{0.17em}in\hspace{0.17em}the\hspace{0.17em}soil}}$$

1

The transition coefficient differs according to the type of crop. The average values for polished rice and potato are 0·0005 and 0·011 respectively (Ochiai et al., 2009).

In the steady state, the general management for two-dimensional seepage (Desai et al., 2011) can be expressed as the equation

$$k_x\frac{{\partial }^{2}H}{\partial x^2}+k_y\frac{{\partial }^{2}H}{\partial y^2}+Q=0$$

2

where H is the total head, k _x is the hydraulic conductivity in the x direction, k _y is the hydraulic conductivity in the y direction and Q is the applied boundary flux. Further, the hydraulic conductivity is estimated to be equal in the x and y directions.

The advection–dispersion equation, which includes the radioactive half-life, can be written as

$$\theta \frac{\partial C}{\partial t}+{\rho }_{\mathrm{d}}\frac{\partial S}{\partial C}\frac{\partial C}{\partial t}=\frac{\partial C}{\partial x_i}\left(\theta {\mathsf{D}}_{ij}\frac{\partial C}{\partial x_j}\right)-U_i\frac{\partial C}{\partial x_i}-\lambda \left(\theta C-S{\rho }_{\mathrm{d}}\right)$$

3

where θ is the volumetric water content, C is the concentration of the radioactive substance, ρ _d is the dry density of the soil,

${\mathsf{D}}_{ij}$

is the dispersion tensor, U _i is the Darcian velocity, t is the time, λ is the decay coefficient and S is the adsorption.

The dispersion tensor

${\mathsf{D}}_{ij}$

The dispersion tensor

${\mathsf{D}}_{ij}$

is shown as the equation

$${\mathsf{D}}_{ij}={\alpha }_{\mathrm{T}}\Vert V\Vert {\delta }_{ij}+\left({\alpha }_{\mathrm{L}}-{\alpha }_{\mathrm{T}}\right)\frac{v_i{v}_j}{\Vert V\Vert }+D^{*}\tau {\delta }_{ij}$$

4

where α _L is the longitudinal dispersivity, α _T is the transverse dispersivity, v _i is seepage velocity, ‖V‖ is the norm of actual flow, D* is the coefficient of molecular diffusion of the radioactive substance, τ is the tortuosity and δ _ij is Kronecker delta (Mohammad, 2013).

Dispersion is a phenomenon where substances spread in the flow direction. Dispersivity is divided into longitudinal and transverse dispersivities. Longitudinal dispersivity is the degree of concentration that is a blurred antecedent to the water velocity in the direction of the water flow. In contrast, transverse dispersivity is the degree of concentration that is blurred towards the direction at right angles to the water flow (Mohammad, 2012a, 2012b), as can be seen in Figure 7. When the analysis scale is approximately 1 m, the vertical dispersivity is determined to be 0·01. In addition, when the analysis scale is approximately 10 m, the longitudinal dispersivity is at 0·1 (National Institute for Environmental Studies, 2012). Moreover, transverse dispersivity is one-tenth of longitudinal dispersivity (Gelhar and Axness, 1983). Advection is a phenomenon of concentration that is mutated by this motion. The molecules move consistently unless they are at the molecular absolute zero point.

Figure 7 Concept of dispersivity. x_n , distance of contaminant diffusion by water velocity when time t passes

The decay coefficient is a parameter for expressing the phenomenon of the concentration decrement as a result of the radioactive half-life. This phenomenon is expressed by an exponential function and can be written as

$$C=C_0{\mathrm{e}}^{-\lambda t}$$

5

where C ₀ is the initial concentration.

The ratio of the concentration to the initial concentration is then 1/2 for the case where the time is the radioactive half-life, T _1/2. This can be seen in the equation

$$\frac{C}{C_0}=\frac{1}{2}={\mathrm{e}}^{-\lambda T_{1/2}}$$

6

This leads to the equation

$$\lambda =\frac{\ln \left(2\cdot 0\right)}{T_{1/2}}$$

7

The adsorption is calculated by using the linear adsorption isotherm with the distribution coefficient K _d, as shown in the equation

$$S=K_{\mathrm{d}}C$$

8

The distribution coefficient is an index for expressing the hydrophobicity and the transitivity of a chemical substance. This index is determined by the radioactive substance and the soil. Moreover, it is the slope of the linear adsorption isotherm.

The measurement results were recorded by Saito et al. (2012) in the Suginoura meeting place, which is approximately 70 km from the Fukushima Daiichi Nuclear Power Plant. The measurements were taken three times, after 9, 18 and 20 months. A scraper plate, the use of which is accepted as the standard sampling method of the depth direction by the International Atomic Energy Agency, as shown in Figure 8, was used for soil sampling. The use of a scraper plate is a method for taking samples by stripping soil off in small amounts from the ground surface and collecting it. Soil samples were taken from the ground surface at depths of 0·5, 0·5–1·0, 1·0–1·5, 1·5–2·0, 2·0–3·0, 3·0–4·0 and 5·0–8·0 cm. Gamma beam spectrum analysis with a germanium semiconductor detector was conducted for measuring caesium-137.

Figure 8 The method of obtaining soil samples by using a scraper plate

Additionally, the soil at this field site was studied by particle size measurement using sieving and laser-diffraction-type equipment to measure the particle size distribution. Sieving was carried out for particle sizes greater than 425 μm. The laser-diffraction-type equipment was used for sizes less than 425 μm. Figure 9 shows the measurement results. According to these results, the concentration at 9 months was distributed exponentially from approximately 10 000 Bq/kg near the ground surface. The concentration near the ground surface reduced by approximately 1000 Bq/kg over 9 months because the concentration at 18 months was almost 9000 Bq/kg. This can be attributed to the four actions described earlier regarding the movement of radioactive substances in the soil. However, caesium-137 remained at a depth of approximately 5 cm for all periods.

Figure 9 Analysis results and the measurement results of caesium-137 concentration distribution in depth direction: (a) 9 months’ progress; (b) 18 months’ progress; (c) 20 months’ progress

Since the decreasing tendency is similar even if time passes, this point might occasionally be disordered by strong rain or factitious external force. Relative to other points, the concentration was distributed in an approximately flat profile until reaching a significant depth. However, for the entire set of measurement results at the field site, the concentration in the depth direction was distributed similarly to an exponential function. The particle size measurement revealed that this point had a particle size of 5–75 μm in the silt layer.

One-dimensional seepage and advection–dispersion analysis, which considered the radioactive half-life, was carried out in order to evaluate quantitatively the concentration distribution in the depth direction in the soil. The section chosen for analysis was the silt layer. Table 1 shows the typical values for silt. The hydraulic conductivity k is obtained from the Japanese Geotechnical Society (2002). The distribution coefficient of silt is based on the work of Inoue and Morisawa (1976). However, the coefficient of molecular diffusion and the initial concentration of caesium-137 at this point were proven by Bossew and Kirchner (2004). The radioactive half-life used for caesium-137 was 30·2 years and the longitudinal dispersivity was assumed to be 0·01 m. These values for caesium-137 are shown in Table 2.

Table 1 Parameters for silt

Table 2 Parameters for caesium-137

Figure 9 shows the measurement and analysis results of the concentration distribution of caesium-137 in the depth direction. The analysed values approach the measured values at each depth. Likewise, the tendency of the decreased concentration in the analysis result is close to the measured result. Furthermore, the values are the same over the passage of time. Therefore, seepage and advection–dispersion analysis, which considered the radioactive half-life, can reproduce the concentration distribution at the field site.

The section for analysis is shown in Figure 10. The transportation behaviours were evaluated at a local section where a difference in water head occurred. Furthermore, the section was saturated so that it was possible to evaluate the most critical situation where caesium-137 had fallen onto the ground surface. The boundaries were coloured blue on a section to represent a drainage condition. The soils for analysis are selected as silt, sand and clay. There are four cases: case 1 is the silt layer; case 2 is the sand layer; case 3 is the clay layer for all ranges of the section used for analysis; and case 4 is alternate layers of silt in section 1 and clay in section 2.

Figure 10 The section for analysis

Typical values are used for each parameter, as shown in Table 3. The hydraulic conductivities k of sand and clay are in accordance with the literature (Japanese Geotechnical Society, 2002). The Radioactive Waste Management Center (1990) studied the distribution coefficients of the sand and clay. Generally, the hydraulic conductivity k of sand is larger than that of clay. The distribution coefficient K _d of sand is smaller than that of clay (Mohammad, 2012a, 2012b). However, caesium-137 was released in large quantities as a result of the accident at Fukushima. Typical values for caesium-137 above the initial concentration and the coefficient of the molecular diffusion at the field site were adopted. Therefore, Table 2 shows the additional transverse dispersivity. Since the height of the section used for analysis was 0·5 m, the longitudinal dispersivity was determined as 0·1 m. The transverse dispersivity was determined as 0·01 m, one-tenth of the longitudinal dispersivity, as shown in Table 4.

Table 3 Parameters for silt, sand and clay

Table 4 Parameters for caesium-137

The streamline vector in the silt layer is shown in Figure 11. The contour lines expressing total head increased from blue to red. According to Figure 11, water movement occurred from areas of large total head to small – namely, from section 1 to section 2. It was particularly strong from the right-hand edge of section 1 to the left-hand edge of section 2, which is confirmed for the other cases. The contours of the concentrations at 1, 2, 5 and 10 years for all cases are shown in Figures 12, 13, 14 and 15, respectively. Caesium-137 is shown to exhibit advection and dispersion along the water flow for all cases.

Figure 11 Flow vector in a silt layer

Figure 12 The transition of caesium-137 in a silt layer after (a) 1 year, (b) 2 years, (c) 5 years and (d) 10 years

Figure 13 The transition of caesium-137 in a sand layer after (a) 1 year, (b) 2 years, (c) 5 years and (d) 10 years

Figure 14 The transition of caesium-137 in a clay layer after (a) 1 year, (b) 2 years, (c) 5 years and (d) 10 years

Figure 15 The transition of caesium-137 in an alternate layer after (a) 1 year, (b) 2 years, (c) 5 years and (d) 10 years

In the silt layer, there was significant movement around the top of the slope. Therefore, caesium-137 reached almost 0·1 m depth after 10 years. However, transportation in the sand layer was larger than that in the silt layer. Caesium-137 reached the ground surface of section 1. Moreover, because caesium-137 leached, the total concentration in the sand layer likely reduced. The result for reaching 0·1 m depth at 1 year is different from the measurement result at the field site. However, the transportation in the clay layer is the smallest of the three. There is a slight movement at the right-hand edge of the ground surface in section 1. The movement was not caused by the hole. Accordingly, the concentration was relatively high at the ground surface.

Transportation of the alternate layer of silt and clay was different from the other cases. Migration was similar to case 2 until 2 years of progress had been recorded in the sand layer of section 1. In practice, caesium-137 does not infiltrate within 9·5 m from the left edge over 10 years. Then, the concentration distribution in the depth direction of the 9·5 m field site was an alternate layer after 10 years of progress, compared with the sand layer, as shown in Figure 16. As Figure 16 indicates, the concentration of the alternate layer was distributed to a depth of 0·2 m. However, the concentration in the sand layer reached a depth of over 0·2 m. It follows that caesium-137 remained in the silt layer since transportation in the silt layer was unlikely to be noticeable, as mentioned earlier.

Figure 16 Concentration distribution at 9·5 m point in an alternate layer and a sand layer

The levels of caesium-137 attained in each period are shown as the concentration distribution in the depth direction of the 0·7 m point, which is the centre of the section used for analysis, as shown in Figure 17. For all sections, the concentration at the ground surface reduced over time. For the silt layer, it was proven that caesium-137 could infiltrate while the concentration was decreasing. However, caesium-137 remained until it reached a depth of 0·05 m from the surface of the ground. In the sand layer, caesium-137 remained at ground level for 1 year. Based on 2 years of progress, the concentration has a shaped waveform. In other words, the most concentrated point appears inside the ground. In contrast, the ground surface was the most concentrated point even after 10 years. The concentration distribution in the alternate layer is a waveform with a time progression similar to that of the sand layer. Although transportation in the alternate layer was smaller than in the sand layer, caesium-137 reached a depth of 0·1 m after 10 years.

Figure 17 Concentration distribution at 0·7 m point in each layer: (a) in a silt layer; (b) in a sand layer; (c) in a clay layer; (d) in an alternate layer

There is a significant difference between the behaviours of radioactive substances in sand and clay layers. The primary factor is an advection term that excels as an adsorption term. In brief, the hydraulic conductivity k has a larger effect on transportation behaviours than the distribution coefficient K _d. When the values of the hydraulic conductivity k and the distribution coefficient K _d of the silt layer shown in Table 3 changed, the transition of the concentration at a point 0·7 m from the left-hand edge and 0·2 m deep was evaluated.

First, the hydraulic conductivity k changed from 1·0 × 10⁻⁴ to 1·0 × 10⁻⁶ m/s for each order, similar to that shown in Figure 18. According to Figure 18, the transition of concentration by changing the hydraulic conductivity k was 1·0 × 10⁻⁴ m/s, and caesium-137 reached the point in earlier years. Then, the concentration gradually fell after the peak at 4 years of progress. When the hydraulic conductivity k was 1·0 × 10⁻⁵ m/s, the concentration increased after 5 years of progress. However, when the hydraulic conductivity k was 1·0 × 10⁻⁶ m/s, the caesium-137 quantity did not reach the point even after 10 years of progress. Thus, significant changes in the transportation behaviour occurred based on the value of the hydraulic conductivity k.

Figure 18 Concentration transition by changing the hydraulic conductivity k

Next, the concentration at the same point was observed when the value of the distribution coefficient K _d was changed from 4·0 × 10⁻³ to 1·0 × 10⁻² g/m³ based on the value of sand compared with that of clay. Figure 19 shows this result. As Figure 19 indicates, the concentration after 10 years decreased as the distribution coefficient K _d reduced. However, there was no difference in the transportation case for the hydraulic conductivity k, and only the concentration ratio increased as the distribution coefficient K _d decreased. For the case where the concentration was the highest, the distribution coefficient K _d was 4·0 × 10⁻³ g/m³, and the increase was approximately 1000 Bq/kg. The influence of the distribution coefficient K _d was smaller than that of the hydraulic conductivity k.

Figure 19 Concentration transition by changing the distribution coefficient K _d

Accordingly, it was confirmed that the advection term excels as an adsorption term. Therefore, the restraint of migration for a sand layer can be expected by making the hydraulic conductivity k small, as in compaction. However, if the ground has a small hydraulic conductivity k, transportation behaviour by dispersion and diffusion needs to be considered.

The influence of radioactive half-life can be evaluated by comparing analysis results that considered radioactive half-life with those that did not. Figure 20 shows the concentration distribution for a sand layer at the 0·7 m point for these two cases. The results show that there is a difference of approximately 2000 Bq/kg at the peak concentration. This difference is important for the environment since decay by radioactive half-life is not affected by an external force. The arrival depth was constant without considering radioactive half-life. Thus, a decay term will not affect transportation in the depth direction.

Figure 20 Difference in concentration distribution between cases where radioactive half-life was considered and not considered

Dredging out a depth of approximately 5 cm as a decontamination method for the analysis section was effective for grounds with silt or clay layers. For ground with an alternate layer, it could be effective for decontaminating a clay layer as it was found that movement would not occur in this layer. Finally, if the ground was a sand layer, a portion of the caesium-137 reached a 0·1 m depth after 1 year. Thus, use of a sand layer is not effective in modern decontamination. Moreover, caesium-137 falling to the ground surface of section 2 after 5 years of progress may be a serious problem. Setting an impervious wall in the ground was proposed as a countermeasure for the sand layer. This is expected to be an effective suppression of transportation as radioactive substances stagnate around an impervious wall.