1. Introduction

The inadvertent release of radioactive materials during the decommissioning of nuclear reactors, the fabrication of nuclear fuel, and the utilization of radioactive materials for industrial applications has the potential to result in contamination of nearby soils, sediments, and groundwater [1,2]. The Fukushima nuclear accident in Japan exemplifies the extent of soil and groundwater contamination and the challenges associated with decontamination of radioactive cesium and strontium, among numerous other radioactive materials [3]. In South Korea, a considerable quantity of soil contaminated with ¹³⁷Cs and ⁶⁰Co was collected and stored in the vicinity of a research reactor site [4,5]. Furthermore, soil and sediment contaminated with depleted ²³⁸U await the development of safe disposal technology. A repository for low- and intermediate-level waste exists in Korea, but in order to extend its lifespan, which is currently estimated at several decades, the volume of the waste must be reduced as much as possible through the application of various waste treatment technologies. Many soil clean-up technologies for radionuclide contaminated soils, including dry separation, flotation separation, Magnetic separation, and wet separation with soil washing were proposed in the many countries.

Dry separation involves excavating radioactively contaminated soil and transporting it on a conveyor belt, while nuclides that emit gamma rays are used to separate contaminated soil from uncontaminated soil in the dry using a gamma scintillation detector such as NaI [6]. The separated uncontaminated soil is used for landfilling, while the contaminated soil is disposed of as radioactive waste. Commercial dry separators, such as Eberlin Services’ Segmented Gate System, were applied to treat soil contaminated with ²³⁹Pu at early U.S. military bases, such as Johnston Atoll DoD [7]. However, when researchers at the Idaho National Engineering and Environmental Laboratory (INEEL) attempted to separate contaminated soil using a segmented gate system to treat soil contaminated with ¹³⁷Cs, they found that the dry separation technique was not applicable because the volume reduction was less than 3% [8]. Dry separation can be applied to inhomogeneously contaminated soils such as excavated soil, but is difficult to apply to homogeneously contaminated soils [9]. Therefore, it may not be effective if the soil is excavated and transported on a conveyor belt, where it is mixed and homogenized, as we are considering here.

In flotation separation, contaminated fine particles are separated from gross soil using techniques based on the surface properties of the particles. Froth-friendly materials are attached to the surface of the bubbles by injecting gas into a dispersed solution or slurry containing fine particles of contaminated soil. The particles attached to the froth float with the froth and are then separated from the top. In a field application, flotation separation technology was studied to separate U from the Elliott Lake ore in Canada [10]. Flotation exploits the hydrophilic/hydrophobic differences in the surface of particles, where uranium and cesium have different properties depending on their chemical bonding state or physical location. In particular, soil’s complex composition, including iron oxides, binds with these radioactive materials, reducing the flotation process’s efficiency. In addition, since flotation works best with high concentrations, it is difficult to separate low concentrations of uranium and cesium, which are typically found in small amounts in soil as is the case we are considering here.

Magnetic separation is a technique to reduce the radioactivity of contaminated soil by separating clay particles containing high Cs concentrations. Various methods have been studied to reduce the volume of radioactive soil; magnetic separation has received particular attention because it is a physical method that does not use adsorbents or chemical reagents (e.g., acids, cations, chelating agents, or surfactants). Other advantages of magnetic separation include simple operation, low energy consumption, and low cost. In field applications, extensive research has been conducted to separate fine particles containing high concentrations of radioactive Cs from real Fukushima soil using modified Fe₃O₄ particles under dry conditions. Fine soil particles coated with magnetic particles are attracted to the magnet against gravity, and the waste is produced in a dry state that is easy to store. Therefore, this method can concentrate radioactive Cs into a smaller volume and reduce the volume of contaminated soil [11]. In fact, the BERAD system has a magnetic separation unit. However, since we are only interested in the desalination performance of particle size separation by water washing, we will not compare the performance of a magnetic separation process here.

Wet separation is typically performed in conjunction with soil washing because radionuclides show the greatest affinity for particles with very high specific surface area/ volume ratios, such as silt and clay. During wet separation with washing, scrubbing operations typically remove surface contaminants from large soil particles such as sand and gravel. Fine soil particles can sometimes be further separated in the settling basin with the help of flocculants. Soil decontamination has been performed at the Cabrera NPP in Spain through aqueous scrubbing [12]. Surfactants, acids, or solvents can be added to the cleaning solution to increase the efficiency of contaminant removal from soil [13]. Soil washing works better on sandy soils with high porosity, good permeability, and adequate moisture content, whereas clay soils or soils with high moisture content are less effective because of their low porosity, low permeability, and the dilution effect of soil water.

The selection of an appropriate treatment technology is dependent upon a number of factors, including the chemical properties of the contaminants, the pathways of contamination, and the physico-chemical characteristics of the soil. BERAD (Beautiful Environmental Construction’s Radioactive Soil Decontamination), one of the soil decontamination technologies, can be combined with dispersion, washing, particle separation, magnetic separation, dewatering, and water treatment units as needed to maximize decontamination efficiency, taking into account the specific contaminants and soil characteristics involved [14,15]. At the BEC facility (Fig. 1), a large soil washing plant constructed according to the same principles has been employed for the separation and disposal of soils and sediments contaminated with a range of hydrocarbons and heavy metals.

Fig. 1

BERAD soil washing system: (a) Top view of the BERAD system; (b) After being washed by the BERAD system, the particle size separation, magnetic separation, and finally the sludge cake are shown.

In general, research on decontamination technology for radiological contaminants has been conducted on a bench scale. This has involved the spiking of stable isotopes into soil samples and the subsequent measurement of the extraction efficiency of the target contaminants in the laboratory. BEC is initiating research by constructing a prototype system, which will be equipped with the requisite safety instruments, including radiation monitors and shielding membranes, for the remediation of soils contaminated with a range of radioactive substances. The success of decontamination with BERAD is contingent upon three conditions: (1) the contaminants must be concentrated in a particular soil particle size fraction, with a sufficiently low volume of the size fraction to reduce the disposal volume of the contaminated fractions; (2) the contaminants must be extractable and removable with a suitable washing solution; and (3) the contaminants must be absorbable by iron-rich minerals that can be separated by an electromagnet removal process.

Objectives of this demonstration was (1) to measure the desired size separation ability of the BERAD system in comparison with laboratory wet sieving method, (2) to measure distribution of indigenous uranium, cesium, strontium, cobalt as surrogate radionuclides among particle size fractions, (3) to measure treatability with unmodified local groundwater. In this demonstration of the washing and size separation process, the effectiveness of the technology was evaluated by monitoring the concentration of naturally occurring uranium, cesium, strontium, and cobalt in a soil sample. It is expected that the physical and chemical behaviors of these elements will be similar to those of radioactive ²³⁸U, ¹³⁷Cs, ⁹⁰Sr, and ⁶⁰Co in contaminated soils and groundwater. Some of these elements can be separated by various extraction solutions, including groundwater. However, they are predominantly adsorbed on soil minerals or present as a constituent element in mineral structures. In the event of an accidental release of radioactive elements, these are primarily absorbed by soils and sediments, as observed in instances of spillage [16]. The behavior and nature of these elements may undergo changes following absorption, contingent on the physicochemical states of the host soils and sediments [17].

2. Experimental Design and Methods

To investigate the removal efficiency of targeted elements in soil by soil washing and size separation, a soil sample weathered from a granite formation in the Goon San City area, was utilized. The soil exhibits a reddishbrown color and contains numerous coarse fragments. The soil pH was neutral with a low content of organic matter. The qualitative mineralogical composition of the soil was determined by X-ray diffraction (XRD). The results of the 0.75 mm to 0.52 mm and <0.52 mm size fractions demonstrated the presence a varying amount of micas, kaolinite, chlorite, hornblende, amorphous iron minerals in conjunction with quartz and feldspar minerals (Fig. 3). The relative intensity of the diffractograms indicated that quartz and feldspars content were increased in coarser fraction, on the other hand, layer silicate minerals were increased in the finer fraction.

The soil was then air-dried at the soil treatment facility while being mixed in order to improve the homogeneity of the soil sample. In the bench scale wet sieving process for the comparison of performance with BERAD system, 150 grams of the air-dried soil was utilized, along with 1.5 liters of groundwater in the laboratory. On the other hand, for the pilot scale BERAD system test 150 kg of the soil sample was washed with 1,500 liters of groundwater for the BERAD system test. The system is requiring minimum such amount of soil and extracting groundwater to pass through the system. The soil was washed with groundwater obtained from the BEC Gam Gok facility. The schematic flow diagram is shown in Fig. 2. The radionuclide-contaminated soil is transported via a screw conveyor, large lumps are shredded in a rotary shredder, and particles larger than 4 mm are filtered out in a trommel screen. Particles smaller than 4 mm washed in the Attrition scrubber and Pressurized Multi-point Injector and collected in the V-bottom tank. The V-bottom tank is equipped with two circular particle segregators, each with a hydro-cyclone on the top, one with sieves of 2 mm and 0.8 mm and the other with sieves of 0.2 mm and 0.075 mm. Therefore, the particles with sizes of 4–2 mm, 2–0.8 mm, 0.8–0.2 mm and 0.2–0.075 mm are separated by the circular particle separator, and the fines smaller than 0.075 mm remaining in the V-bottom tank by the second hydro-cyclone are finally made into sludge cake by the filter-press.

Fig. 2

Schematic flow diagram of BERAD soil washing system.

To reduce the potential error in the leachate composition, the 50 grams of soil were washed with 500 milliliters of the groundwater (Table 1). The composition of the groundwater, extracted waters, and washed soil size fractions was determined using a range of standard water and soil analysis methods, as well as ICP-MS and ICP-AES methods following acid dissolution, a process known as Aqua Regia Extraction [17,18]. As shown in Table 1, the pH of the sample soil is 8.0 and the groundwater used for washing is 7.8. After washing the soil, the pH of the extraction water slightly increased to 7.9. For the radionuclides of interest in this experiment, U, Sr, Co, and Cs, the amount dissolved in the water was very small, and the nuclides in the groundwater were adsorbed onto the soil (see the negative sign in the “Differences” row). Therefore, the leaching effect of water was not a concern. The mineralogical composition of the soil was determined by X-ray diffraction (XRD). The results of the 0.75 mm to 0.52 mm and <0.52 mm size fractions demonstrated the presence a varying amount of micas, kaolinite, chlorite, horn-blende, amorphous iron minerals in conjunction with quartz and feldspar minerals (Fig. 3).

Fig. 3

X-ray diffractograms of (a) 0.075 mm–0.052 mm and (b) <0.052 mm size fractions of the soil.

3. Results and Discussion

Soil particle size separation: Particle size separation was conducted via wet sieving and the BERAD system, and the resulting data are presented in Table 2. Wet sieving, performed by hand with a 500 g soil sample and sieves identical to the sieve size of the BERAD system, was performed to compare the particle size separation performance of the BERAD system. As particle size decreases, the weight of the size fractions increases during wet sieving. The size distribution of the BERAD system exhibits a comparable trend, although the quantity of the <0.075 mm fraction is less than that of the 0.2 mm−0.075 mm size fraction. This indicates that the hydro-cyclone setting for particles with a diameter of less than 0.075 mm was smaller than the targeted particle size. The particle size separation results demonstrate the efficiency of the BERAD system and illustrate its flexibility in terms of setting desired size separation parameters as required. The BERAD system could be employed for the remediation of soils contaminated primarily with particles within specified size ranges by establishing upper and lower size limits during the soil washing process. The size segregation approach has been demonstrated to be an effective method for the remediation of soils contaminated by airborne particles, as evidenced by the experiences at the Nevada Test Site and the Fernald Environmental Management Project Site [19,20]. In general, radionuclides released through bomb test and incinerator operation surrounding area soil were remained in particulate form. The particle size of contaminants related distance from the source. If one would find dominant size of contaminants, the BERAD system could separate the desired size fraction by adjusting sieve size and speed of hydro cyclones in the system.

The results of particle size separation and washing of the contaminated soils are an important criterion for the design of decontamination approaches. While some of the contaminated radioactive materials can be removed by a specially designed extraction solution, some of them are insoluble or irreversibly adsorbed on fine soil particles. Consequently, if the amounts of heavily contaminated particles are a small fraction of the total soil volume, the removal and disposal of the fraction could reduce the volume of the wastes and the cost of disposal.

The origin and the characteristics of naturally occurring U, Sr, Co, and Cs in soil: Uranium is a radioactive element that occurs in trace amounts in soil [21]. Given its utilization as nuclear fuel, the element has manifested as a contaminant in soil and groundwater. Background concentrations of uranium in soils and groundwater vary according to the geologic formations. This study examined the potential of utilizing naturally occurring uranium in soil as an alternative to contaminated enriched or depleted uranium in the event of an accidental release. In the event of an accidental release of uranium-containing solutions or particles into the soil, the absence of immediate removal treatment may result in the absorption of uranium by iron oxides and layer silicate minerals, leading to a change in state that renders it immobile, akin to its natural state. Accordingly, to assess the efficacy of soil washing, the indigenous uranium present in the soil was employed in lieu of uranium in the soil that had been inadvertently contaminated.

Strontium is a minor element in various carbonate and phosphate minerals, as well as being adsorbed as a cation on the surface of oxides and layer silicate minerals. Radioactive ⁹⁰Sr is typically produced by the fission of uranium, and may be accompanied by the release of radioactive ¹³⁷Cs. The chemical characteristic of radioactive ⁹⁰Sr released by an accident may vary depending on the chemical composition of the contaminated water and the mineral composition of the hosting soils [22].

Cobalt, a heavy metal, is present in trace amounts in various minerals, including other heavy metals in soil. Some of the dissolved cobalt may be adsorbed on the surface of oxidized minerals (iron, aluminum, manganese oxides) and clay minerals in soil, while others migrate through groundwater. Radioactive cobalt (⁶⁰Co) utilized as a gamma-ray source is produced through a neutron activation process of non-radioactive ⁵⁹Co. However, if released accidently in a soluble form, it has the potential to contaminate soils and ground water. Accordingly, this study assessed the decontamination efficacy of radioactive ⁶⁰Co in a hypothetical soil contamination scenario through soil washing and particle size separation utilizing naturally occurring ⁵⁹Co in the soil.

Cesium is a common alkali element in naturally occurring minerals, particularly those containing potassium (K), which form part of the mineral structure and are absorbed by layer silicate minerals. The ¹³⁷Cs produced from ²³⁵U fission is water-soluble and, like natural cesium, when it comes into contact with soil minerals, it is absorbed by layer silicate mineral particles, and cannot be easily removed by simple extraction methods [23,24].

Concentration Distribution of U, Sr, Co and Cs from experiment: Following the wet sieving separation, the concentration of uranium was found to be highest in particles of <0.075 mm size (Table 3). The weight of these particles constituted 27.2% of the total soil and accounted for 35.2% of the total uranium content of the soil. The concentration of uranium following separation via the BERAD system was also highest in the 0.2–0.075 mm and <0.075 mm size fractions (Table 3). In conclusion, 35.8% of the total soil was composed of particles <0.2 mm in size, and the amount of uranium was 53.2%. In a separate batch washing experiment, uranium content in extracted groundwater was decreased from 0.005 ppm to 0.002 ppm, showing adsorption of dissolved uranium on soil minerals which have 4.3 ppm uranium as indigenous components. The retention of the dissolved uranium was attributed by the high iron content in the soil. Iron oxide and coating in soil minerals have a high adsorption capacity for uranium as well as other radionuclides such as Co, Cs, and Sr. The high correlation between these elements and iron concentrations in various size fractions, albeit to varying degrees, is shown in Fig. 4. The distribution of iron and uranium content in the varying size fractions in this experiment exhibits a high correlation (Fig. 4(a)). The presence of iron in soil is amorphous and can be adsorbed onto soil particles or present as iron oxide minerals (e.g., hematite and goethite), which can adsorb a range of radioactive substances, including uranium.

Fig. 4

Illustrates the correlation between uranium, strontium, cobalt, cesium and iron content in the size fractions.

The distribution of the natural background strontium content in the soil exhibited an increase with a decrease in particle size, and a similar trend was observed for the iron content (Fig. 4(b)). The results of the particle size fractionation conducted using the BERAD method, along with the distribution of strontium within the particles, indicate that the fraction comprising particles with a diameter of less than 0.2 mm accounts for 35.8% of the total weight and contains 45.3% of the total strontium present within all size fractions (Table 3). It is likely that strontium dissolved from primary minerals during weathering could be immobilized on the surface of iron and aluminum oxides, as well as layer silicate minerals (kaolinite, micaceous minerals).

The BERAD and wet sieve particle size separation demonstrate that the cobalt content increases with decreasing particle size (Table 2). The results demonstrate that the increase in cobalt is accompanied by an increase in iron content across various size fractions (Fig. 4(c)). Additionally, the XRD results indicate that a significant portion of cobalt is adsorbed on iron oxide and layer silicates. The results of the particle size fractionation conducted using the BERAD method, along with the distribution of cobalt within the particles, indicate that the fraction comprising particles with a diameter of less than 0.2 mm accounts for 35.8% of the total weight and contains 66.0% of the total cobalt present within all size fractions (Table 3).

The distribution of cesium demonstrates an increasing trend with decreasing particle size, particularly for particles <0.075 mm, as observed in both washing methods. The distribution of cesium exhibits a similar trend to that of iron (Fig. 4(d)). The results of the particle size fractionation conducted using the BERAD method, along with the distribution of cesium within the particles, indicate that the fraction comprising particles with a diameter of <0.2 mm accounts for 35.8% of the total weight and contains 52.8% of the total cesium present within all size fractions (Table 3). This indicates that the <0.075 mm particles contain a substantial quantity of micaceous minerals and iron oxides, as evidenced by the XRD results. These particles possess high selectivity for cesium adsorption, a process that is almost irreversible. It is established that cesium is immobilized on the surface of cation exchange layer sites of micaceous minerals, rendering it difficult to substitute with other cations [20].

Discussion: As evidenced by the groundwater elution results, the efficacy of uranium, cobalt, strontium, and cesium decontamination is significantly diminished in the absence of an appropriate eluent for each contaminating radionuclide (Table 1). The pH of groundwater (7.8) was very similar to pH of the soil (7.9) and there were no drastic changes in chemical equilibrium states during elution experiment. In the event that the soil originates from an area that has been contaminated, the elution rate is anticipated to be higher than that observed in the soil that has been investigated in this study. Furthermore, this would be dependent upon the extent of mineralization of the radioactive substances present in the contaminated soil.

In this study, the concentration of naturally occurring background elements in the soil was sufficient to indirectly demonstrate the potential for waste reduction through particle size separation and washing in place of their radioactive counterparts. In this soil washing demonstration, it can be assumed that if the <0.2 mm particle fraction is removed during the BERAD washing process, the 53.2% of U, 45.3% of Sr, 65.9% of Co, and 52.8% of Cs can be removed. The concentrations of naturally occurring background elements and their corresponding regulatory reference doses were compared in terms of concentration (Table 1). The findings of this study indicate that future soil decontamination experiments may be conducted without the inclusion of radioactive elements to evaluate removal efficiency.

4. Conclusion

The BERAD system was developed with the objective of reducing the volume of radioactive waste by means of cleaning and separating highly contaminated particle size fractions from the radioisotope-contaminated soils. Its performance was tested using indigenous stable isotopes such as uranium (U), cobalt (Co), strontium (Sr), and cesium (Cs) in natural soils. Results showed that the concentrations of U, Sr, Co, Cs, and Fe increased with decreasing particle size, and it can be assumed that if the <0.2 mm particle fraction is removed during the BERAD washing process, the 53.2% of U, 45.3% of Sr, 65.9% of Co, and 52.8% of Cs can be removed. These results show that if radionuclides are contaminating the soil, even if the experiment was performed with non-radioactive isotopes, the separation of fine soil by the soil washing method is effective in removing radionuclides. This experiment was conducted with natural soil from a smelter, which contains a lot of iron and most of the nuclides are adsorbed on iron, but in the case of an accident at a nuclear power plant where radionuclides are artificially contaminated, the effect of soil washing may be better if most of the Cs particles are distributed in the fine soil. According to the Nuclear Safety Act, self-disposal is possible if the nuclide-specific concentration is found to be below the permissible concentration for self-disposal or if the annual individual dose is lower than 0.01 mSv·y⁻¹, and the total collective dose is lower than 1 person-Sv·y⁻¹, concurrently [25]. Although the tests did not use radionuclides, it is expected that the soil decontaminated by the BERAD wash will be self-disposing, which would reduce the amount of contaminated soil and thus the amount of waste that would need to be sent to a radioactive waste disposal site. If a clearance level is not reached with a single water washing, additional decontamination methods, such as acid washing, may be used to achieve the goal of decontamination of radioactively contaminated soil. Additionally, in instances where the decontamination effect is constrained by particle size separation, it is imperative to identify a cost-effective washing solution that considers the characteristics of the contaminants and the host soils. Therefore, the efficacy of the washing solutions in removing contaminants should also be evaluated during the technology assessments. The results of this study demonstrate that the concentration of naturally occurring uranium, cobalt, strontium, and cesium in soil samples can be accurately quantified using standard analytical techniques, even after washing and particle size separation. The same approach can be applied to a range of solvent extraction experiments in the development of soil washing technology in the future.