Cost and Effectiveness of Decontamination Strategies in Radiation Contaminated Areas in Fukushima in Regard to External Radiation Dose

The objective of the present study is to evaluate the cost and effectiveness of decontamination strategies in the special decontamination areas in Fukushima in regard to external radiation dose. A geographical information system (GIS) was used to relate the predicted external dose in the affected areas to the number of potential inhabitants and the land use in the areas. A comprehensive review of the costs of various decontamination methods was conducted as part of the analysis. The results indicate that aerial decontamination in the special decontamination areas in Fukushima would be effective for reducing the air dose rate to the target level in a short period of time in some but not all of the areas. In a standard scenario, analysis of cost and effectiveness suggests that decontamination costs for agricultural areas account for approximately 80% of the total decontamination cost, of which approximately 60% is associated with storage. In addition, the costs of decontamination per person per unit area are estimated to vary greatly. Appropriate selection of decontamination methods may significantly decrease decontamination costs, allowing more meaningful decontamination in terms of the limited budget. Our analysis can help in examining the prioritization of decontamination areas from the viewpoints of cost and effectiveness in reducing the external dose. Decontamination strategies should be determined according to air dose rates and future land-use plans.


Introduction
The accident at Tokyo Electric Power Company's (TEPCO) Fukushima Dai-ichi Nuclear Power Plant (hereinafter referred to as F1NPP) in March 2011 released radionuclides into the environment and contaminated large areas of land in Japan. Approximately 22% of radionuclides released are thought to have been deposited on land [1]. Monitoring of large areas by aircraft has revealed a zone of high-density surface deposition of radionuclides (primarily caesium 134 and caesium 137) extending to the northwest from the nuclear power station [2]. The contaminated land area in Fukushima Prefecture with a potential annual air dose of 5 mSv extends for 1,778 km 2 [3]. Due to concerns about the possibility of a large-scale release of radioactive materials and health risks, areas within a 20-km radius of the F1NPP (Restricted Area) and heavily contaminated areas outside of this zone (Deliberate Evacuation Area) have been designated. In days and weeks following the accident, approximately 85,000 people from 11 municipalities were evacuated from these evacuation areas, which cover approximately 1,170 km 2 [3]. As of November 30, 2012, the Restricted Area and the Deliberate Evacuation Area have been rearranged into three areas: Area 1 (in which evacuation orders are ready to be lifted), Area 2 (in which residents are not permitted to live), and Area 3 (in which residents will face difficulties in returning for a long time) [4].
Following the accident, the Japanese government promulgated the Act on Special Measures Concerning the Handling of Environment Pollution by Radioactive Materials Discharged by the NPS Accident Associated with the Tohoku District: Off the Pacific Ocean Earthquake That Occurred on March 11 (Act on Special Measures Concerning the Handling of Pollution by Radioactive Materials) on December 8, 2011, with the goal of quickly reducing the impact of environmental pollution from the radioactive material on human health and the environment. Under this legislation, which was enacted January 1, 2012, a framework and guidelines for decontamination operations were released as Decontamination Guidelines (December 2011) [5], which covered methods for surveying and measuring the degree of contamination of the environment in intensely contaminated areas, as well as measures for decontamination and guidelines for collection, transport, and storage of removed soil. The Ministry of Environment also formulated a decontamination plan to be implemented under the direct supervision of the government and announced the Policy for Decontamination in the Special Decontamination Areas (Decontamination Road map; January 2012) [6].
Due to the societal interests and regulatory needs, numerous radiation monitoring or modelling studies to grasp and analyse the current situation in Fukushima have been conducted [7][8][9]. Pilot-scale investigations have also been conducted in order to develop efficient decontamination techniques and to evaluate their respective decontamination efficiencies [10,11]. These pilot-scale investigations revealed that the decontamination efficiency varies depending on land use or air dose rate, and even if decontamination is performed, the air dose rate cannot be reduced to the background radiation levels [12]. Most of the previous studies or reports regarding radiation contamination in Fukushima were rather reductionistic and focused primarily on determining the magnitude of contamination and the development or improvement of decontamination technologies. In order to establish effective and pragmatic decontamination methods for use in the radiation contaminated areas in Fukushima (primarily in the special decontamination areas), a holistic approach for assessing decontamination strategies, their costs, and long-term external radiation doses is needed. For anthropogenic pollution, the polluter pays principle (PPP), in which the polluter is responsible for decontaminating and restoring the affected areas to their original states, should be applied. However, as mentioned above, the special decontamination areas alone span more than 1,100 km 2 , and the effectiveness of decontamination is limited, so that it is impossible to completely restore this area to its original state through decontamination in a short period of time.
Several studies have evaluated the effective dose and the effect of remediation of radioactive contamination in the case of the Chernobyl accident [13][14][15][16]. Most of these studies are retrospective, and analyses on the effects of decontamination were rather confirmatory. Yasutaka et al. [17] conducted a prospective quantitative evaluation of the reduction of the cumulated effective doses due to external radiation exposure for several decontamination scenarios using geographical information system (GIS) in special decontamination areas of Fukushima. However, the costs of the decontamination scenarios were not included in their analysis. Despite the enormous cost associated with radiation decontamination, almost no quantitative assessment has been performed on the relationship between the potential reduction in long-term radiation exposure and the costs of the various decontamination strategies considered for the special decontamination areas in Fukushima. There have been no studies to evaluate the efficiency of decontamination in a timely manner that consider the costs of multiple decontamination techniques.
The primary goal of our study is to provide quantitative information that is useful for decision-making in determining the prioritization of decontamination in the special decontamination area in Fukushima. The objective of the present study is evaluation of the costs and effectiveness of decontamination strategies in the special decontamination areas in Fukushima in regard to external radiation dose. GIS was used to relate the predicted external dose in the affected areas to the number of potential inhabitants and the land use in these areas. The choice of decontamination strategies or countermeasures in the special decontamination areas should be based on a comprehensive analysis of multiple attributes such as radiological, economic, and socio-psychological attributes. The cost and effectiveness of the different decontamination strategies is not the determinant of the remediation strategies of the special decontamination area but is one of the most important attributes when making the policy decision. In the present study, we focus on radiological and economic attributes in determining decontamination strategies.

Target area
The present study considers the contaminated areas from which people have been relocated after the accident. The target area of the present study is the special decontamination areas, where decontamination is being implemented by the national government. For this category, the "Restricted Area" (the area within a 20-km radius of the power station) and the "Deliberate Evacuation Area" (the area in which the annual cumulative dose could exceed 20 mSv) were designated in parts of 11 municipalities in eastern Fukushima Prefecture ( Figure 1). Approximately 86,000 individuals were living in the study area as of year 2010 [18] (Table 1).

Air Dose Rate, Population Density, and Land Use
GIS is a powerful data integration and spatial analysis tool. In the present study, ArcGIS ver. 10.1 is used to aggregate, synthesize, and analyse large datasets and to identify spatial relationships among air dose rate, population density, and land use in the target area. Data on air dose rate (µSv/h) as of November 5, 2011 is obtained from MEXT [19] and the inverse distance weighted (IDW) approach was used to generate a continuous map. The data cover a range of within 80 km of the F1NPP and indicate the air dose rate distribution at a height of 1 m above the ground. For the decontamination efficiency analyses, 100-m-mesh data on the air dose rate were created using a continuous map. The air dose rate at the centre of each of the 100-m meshes was used as the representative value, and the arithmetic mean of the 100-m-mesh air dose rates was then used to calculate the air dose rate for each 1-km-mesh unit (mesh size: 1 km × 1 km). These values included a background level of radiation (average: 0.04-0.05 µSv/h) [20].
The 1-km-mesh population density data for the target area were obtained from the 2010 Population Census 2010 [18]. For the land-use data, we used 100-m-mesh data from national land numerical data provided by the Ministry of Land, Infrastructure, Transport and Tourism [21]. We reorganized the original 12 types of land use into six types: paddy fields, other agricultural land, forest, land for buildings, roads, and other uses. In the following analysis, we consider paddy field and other agricultural land as agricultural areas, land for buildings as residential and building areas, roads as road and street areas. We do not consider other uses (e.g., lakes and rivers, waste lands) as the subject of decontamination.
For the total road length, we used the road density and total road length mesh data from the national geographical numerical data [21]. These data comprise the total extended road distance by road width for each 1-km mesh.

Estimation of external dose
Radiation exposure may be internal or external and can be acquired through various exposure pathways. Dose estimation models that consider both types of pathways can be found elsewhere [14,15]. Since the external dose due to deposited radionuclides is a main pathway of the total radiation and the classification of decontamination areas is based solely on the additional external dose [22], only external exposure is considered in the present study.
The additional annual external dose, D ext,i [mSv], at location i, is calculated as follows: where p i is the air dose rate [µSv/h] at location i, p nat is the air dose rate [µSv/h] from naturally occurring radionuclides, T out is time spent outdoors [hours/day] at location i, T in is the time spent indoors [hours/day] at location i, and β is the shielding factor, which is generally expressed as the ratio of the on-site external radiation level indoors to the on-site external radiation level outdoors. This simple equation is used by the central government of Japan to classify decontamination areas [22]  under the Act on Special Measures Concerning the Handling of Pollution by Radioactive Materials. In order to estimate the annual external dose, the value of p nat was found to be 0.04 [µSv/h], the indoor T in and outdoor T out dwell times were set to 16 hours and 8 hours, respectively, and the shielding factor was set as 0.4. These are the same values used by the central government [22]. In a very simple expression, the additional dose from external irradiation is calculated by the absorbed dose in air multiplied by a reduction factor of 0.6. Evidence has indicated that the additional dose from external irradiation may be well below the estimated additional dose calculated by the simple model [11]. Realistic dose estimations from external irradiation exposure can be an important issue. Since a full discussion of this issue is beyond the scope of the present study, in order to be consistent with the existing regulatory approach, we followed the assumption of the central government in the present study, unless otherwise stated.

Decontamination methods, efficiency, and cost
Appropriate selection of decontamination strategies during the rehabilitation stage of radiation-affected areas can significantly affect decontamination costs. The costs of decontamination, which depends on the land use of the areas, can be classified as 1) clean-up costs and 2) storage costs. The clean-up costs include the cost of removal or reduction for contaminated materials and soil, containers for temporary storage, and temporary storage, while the storage costs include the cost of volume reduction, interim storage, and final disposal. Since no information is available regarding the cost of final disposal, this cost was not considered in the present study. The decontamination methods and their respective unit costs for the various land uses considered in the present study are presented in Table 2 and Table 3. Cost data for various decontamination methods were based primarily on data from previous studies [10,11]. A comprehensive report was published by JAEA [11] on different decontamination methods based on a large-scale pilot study conducted in Fukushima in 2011 and 2012. Estimating the cost of interim storage is difficult because there has been no decision as to whether to implement volume reduction or where or how to go about volume reduction of the contaminated waste and soil generated due to clean up. In order to estimate the magnitude of the interim storage cost, however, we have attempted to estimate the cost of interim storage based on the unit cost of the existing controlled or shielded landfill. Combustible waste was assumed to subject to volume reduction. Moreover, the indirect costs and the costs of radiation control were not considered in the present study. Due to a lack of information, the general administrative expenses and the cost related to decreasing working efficiency in high-air-dose-rate areas were not considered in the present study.
The efficiency of decontamination depends on the air dose rate and land use [17]. The values of decontamination efficiency (φ) for different land use and air dose rates considered in the present study are shown in Table 4. No data are available for the decontamination efficiencies for the air

Results and Discussion
Spatial relationships among air dose rate, population density, and land use with or without decontamination Figure 2 shows the spatial relationships among the estimated air dose rate, potential population density, and land use with or without decontamination in the special decontamination areas as of April 1, 2014 for SC1. Approximately more than 70% of land is covered by forest, and coastal regions have high population density. Assuming that the forest in the vicinity of the residential area is the target of decontamination, the potential decontamination area of the present study covers an area of approximately 437 km 2 . According to the official estimation equation of external dose, within this potential decontamination area, the areas in which the annual external dose is estimated to exceed 1 mSv (air dose rate: approximately 0.19 µSv/h), 5 mSv (air dose rate: approximately 0.95 µSv/h), and 20 mSv (air dose rate: approximately 3.8 µSv/h) on April 1, 2014 would cover 277 km 2 , 196 km 2 , and 55 km 2 , respectively. Without decontamination, the area exceeding an annual external dose of 1 mSv was estimated to cover 288 km 2 , the area exceeding an annual external dose of 5 mSv was estimated to cover 222 km 2 , and the area exceeding an annual external dose of 20 mSv was estimated to cover 94 km 2 . With decontamination, approximately 2,300, 28,900, and 73,100 potential inhabitants were estimated to be returnable for the target levels of 1 mSv, 5 mSv, and 20 mSv, respectively (Figure 3). Approximately 30,000 potential inhabitants would live in areas of relatively low contamination with annual external doses of less than 5 mSv with decontamination. The 1-km-mesh of the highest population density, which is located in Tomioka Town, would have estimated annual external doses of 25 mSv (air dose rate: approximately 4.8 µSv/h) without decontamination and 14 mSv (air dose rate: approximately 2.7 µSv/h) with decontamination on April 1, 2014. These results indicate that aerial decontamination in the special decontamination area would be  Cost Effectiveness of Decontamination in Fukushima PLOS ONE | www.plosone.org effective in reducing the air dose rate to the target levels for some but not all of the areas. A reduction factor of 0.6 was used in the official equation for estimating external dose [22]. External doses estimated using a reduction factor of 0.6 may be conservative and may overestimate the external effective doses in the affected area. In the external dose estimation at Chernobyl, the reduction factors ranged from 0.20 to 0.36, depending on the ages and occupations of individuals [25]. Yasutaka et al. [17] revealed that the selection of reduction factor in estimating external doses is important to determine the area of decontamination. In the post-emergency phase, as knowledge and information regarding external dose in Fukushima have been accumulated, careful consideration of the reduction factor should be incorporated into the examination of future remediation plans of the affected areas.

Estimated cost of decontamination in the special decontamination areas
The estimated costs of different decontamination methods for various land uses and the estimated costs for different decontamination scenarios in the special decontamination areas in Fukushima are shown in Table 5 and Table 6, respectively. Assuming the standard decontamination scenario, SC1, the estimated total cost of decontamination is approximately 1,300 billion JPY. Based on the estimation result, the decontamination costs for agricultural land were estimated to be more than 1,000 billion JPY, which account for approximately 80% of the total decontamination cost. Of the total cost, the percentages of decontamination costs in SC1 are estimated to be approximately 9% for forest areas, 8% for residential areas, and less than 1% for roads and streets. Of the total decontamination cost for agricultural land, the cost of storage is estimated to be 60%. For SC2 and SC3, the estimated cost of interim storage varies widely depending on the requirements of the interim storage. As an alternative to soil removal, which requires storage of a large volume of removed soil, either deep plowing or reversal tillage can be used to decontaminate agricultural land, making it possible to significantly reduce the cost of removed soil storage. According to the cost estimations for SC4, SC5, and SC6, the cost of decontamination in agricultural land would decrease by 99% in the case of using reversal tillage. Our calculation suggests that the appropriate selection of a decontamination method would significantly decrease the cost of decontamination. Note that if the plough pan is shallow, ploughing or reversal tillage cannot be applied because it may cause subsurface leakage of water, resulting in useless paddy fields. Assuming full decontamination of the forest areas (SC7), which seems unlikely, the estimated cost of decontamination would rise significantly.
For decontamination efforts under the Act on Special Measures Concerning the Handling of Pollution by Radioactive Materials, more than 1,500 billion JPY has been set aside to date [26] and additional budgetary requests are expected. The rationale of the proposed costs by the central government is unclear. Compared with the cost of the decontamination effort proposed by the government, the results of the presents study indicate that the proposed costs are underestimated. The cost of decontamination would increase significantly if the Intensive Contamination Survey Area is included in the estimation. In addition, the IAEA [24] criticized the plan to remove large volumes of low-radiation topsoil and waste for storage at secure facilities for an extended period. According to the IAEA [24], this could create unnecessary major challenges without  providing any benefit in terms of reducing the radiation dosage. Even so, the total cost of decontamination would vary significantly depending on which decontamination strategy is selected, especially in the case of agricultural land. The costs of decontamination per person per unit area (1 km 2 ) were estimated for SC1 (Figure 4). The cost was estimated to range between 10 thousand JPY/person and 1 billion JPY/person. The geometric mean of the costs of decontamination per person per unit area was estimated to be approximately 28 million JPY. Approximately 14% of unit areas were estimated to require greater than 100 million JPY/person to conduct decontamination. Our analysis suggests that the cost of decontamination per person for each unit area varies greatly depending on population density and the type of land use. Compared with urban areas having high population densities, agricultural areas would have higher decontamination costs per person. If the objective of decontamination is to maximize population living in the affected areas after decontamination, which could be a cost-effective strategy, then special consideration should be given to the decontamination of urban residential areas. As shown in the previous section, assuming SC1 is implemented, the decontamination costs for agricultural areas are high and the efficiency of the decontamination is limited to a certain degree. Although restoring the original state of an area by decontamination can be a basic philosophy for large-scale anthropogenic pollution, it is important to understand the limitations of current decontamination techniques. As such, the decontamination strategies and areas should be prioritised in terms of cost, effectiveness, land use, and radiological protection (external dose exposure) after decontamination. Complete decontamination of the area cannot be the ultimate goal, and decontamination strategies should be examined considering the rehabilitation planning of the affected areas. However, land use for food production and the resulting internal doses, which were not considered herein, may modify this analysis. Figure 5 shows the trends of potential inhabitants returning to their former residence areas for different dose levels with and without decontamination. According to prospective calculation, in the year 2019, the numbers of the potential inhabitants with annual doses below 1 mSv, 5 mSv, and 20 mSv increase to approximately 11,600, 40,300, and 84,300, respectively with decontamination. In the year 2029, the number of potential inhabitants with annual doses below 1 mSv, 5 mSv, and 20 mSv are estimated to be 19,500, 60,500, and 85,900, respectively.

Long-term trends in the potential inhabitants and areas after decontamination
The majority of the potential inhabitants are estimated to be able to return to their former areas of residence 5 years after the completion of the aerial decontamination if the target level is set to 20 mSv/y and the public attitude toward returning home is ignored. If the target level is set to 5 mSv/y, a distinct difference in the number of potential inhabitants with and without decontamination was predicted after several years of the proposed completion of decontamination. The figure also shows that setting the target level to 1 mSv/y results in a small difference between the number of potential inhabitants with and without decontamination, and approximately 20,000 potential inhabitants are expected to live in the affected areas, even 30 years after the proposed completion of decontamination. Note that all of the predictions of external dose in the present study were based on the results of the Fourth Airborne Monitoring Survey by MEXT [19] assuming only natural radioactive decay and ignoring realistic occupancy/shielding factors. The prediction may be conservative and may overestimate the actual external dose in the affected area. In addition, we used the demographical statistics for 2010 to estimate the number of potential inhabitants returning to their former residences. This tends to overestimate the number of potential inhabitants returning of their former residence. Many of the former residence may not return to their former homes even after decontamination.