The focus of this chapter is on offsite emergency responses to the Fukushima Daiichi accident and lessons learned for emergency preparedness in the United States. The chapter focuses primarily on offsite responses during the first few critical days of the accident (early phase, see Sidebar 6.1). However, information about the early-phase response was useful to the committee for identifying lessons learned for the intermediate and late (or recovery) phases.
The information used in this chapter was obtained from several sources: independent examinations of the Fukushima Daiichi accident carried out in Japan, the United States, and other countries (see Table 1.1 in Chapter 1); Japanese regulations related to offsite emergency management in Japan at the time of the accident; and a number of scientific publications. The committee’s review of Japanese documents was limited to those translated to English. At the committee’s request, the National Academy of Sciences arranged for English translations of selected sections of the Japanese government’s 2007 version of the “Basic Plan for Emergency Preparedness” (NSC, 2013).
At the time of the March 2011 Fukushima Daiichi accident, Japanese1 and U.S. approaches to offsite emergency response had many common features. These included specified incident notification levels; guidance on conditions for each notification level; designation of specific emergency planning zones; protective action guidelines (PAGs) for decisions relating to
1 Described in NSC (2013).
During a radiological emergency in the United States, there is a generic framework for structuring responses following a disaster based on three phases: early, intermediate, and late.
According to the USEPA (2013), the early phase (also referred to as the emergency phase) lasts from several hours to several days. During this phase, conditions at the location of the incident are evaluated, responsible authorities are notified, and the potential consequences of the incident to members of the public are predicted or evaluated. Decisions on protective actions such as evacuation, sheltering in place, and taking KI for thyroid protection are made primarily on the basis of the status of the nuclear power plant and the prognosis of changes in the conditions.
The intermediate phase lasts from weeks to months. During this phase, the source and releases from the plant have been brought under control. Also, environmental measurements of radioactivity and dose models are available to project doses to members of the public and base decisions on additional protective actions such as food and water interdictions.
The late phase (also referred to as the recovery phase) can last from months to years. It begins sometime after the initiation of the intermediate phase and proceeds independently of the protective actions implemented during that phase. During the late phase, recovery actions designed to reduce radiation levels in the environment are commenced and end when all recovery actions have been completed.
Because of the possible overlap, phases of the emergency response are not viewed in terms of time but instead in terms of activities performed.
shelter, evacuation, and distribution and administration of potassium iodide (KI)2; and guidelines for food and water intake.
However, approaches to managing offsite responses in Japan and the United States were different in some notable ways: the United States uses a “bottom-up” approach to managing offsite emergency response. That is, the responsibility for responding to a disaster begins at the local level, extends to state and tribal governments, and can include the federal government as supplemental resources are requested (Sidebar 6.2).3 The Japanese
2 KI is a prophylactic agent that prevents the uptake of radioactive iodine (i.e., radioiodine) into the thyroid gland and thus reduces the risk of thyroid cancer.
3 There are limited exceptions to this approach. The President of the United States is authorized to support precautionary evacuation measures, accelerate federal emergency response and recovery aid, and provide expedited federal assistance (coordinated with the state to the extent possible) in the absence of a specific request from state officials (Robert T. Stafford
The roles of federal agencies in U.S. nuclear emergencies are laid out in the National Response Framework (NRF) (USDHS, 2013). The NRF creates a broad-based “all-hazards response” emergency planning process to address a wide variety of emergencies including natural disasters, terrorism, and other human-initiated accidents and events, encompassing nuclear and radiological events. Nuclear power plant accidents involving radioactive material releases are just one of the many potential emergencies to which this all-hazards approach applies.
The all-hazards approach is based on the notion that there are common features among disasters irrespective of their initiating events; therefore, many of the same planning strategies can apply to all emergencies. These features include the need for robust communication channels; collection of adequate data; information exchange and interpretation; requisitioning of resources and expertise; assessment and management of offsite impacts; and community involvement. Many elements necessary to an effective response to a nuclear incident are common to other types of emergencies, such as sheltering or evacuating a specific population, establishing an emergency communication network, or implementing mutual-aid agreements with nearby (but unaffected) jurisdictions.
Thus, embedding planning for a nuclear-related event in an overall emergency response plan for all types of natural and man-made emergencies provides the framework for a scalable, flexible, and adaptable plan that is expected to be responsive to small, common, and well-defined events as well as large, rare, and complex events. An additional advantage of the all-hazards approach is that it maintains a higher state of readiness, because the plan is implemented more often and because all response agencies, nongovernmental organizations, and private entities are working within the same response framework with a common command structure.
approach at the time of the Fukushima Daiichi accident was “top down,” with the central government providing direction and national resources to local communities (NSC, 2013). Despite these differences in approaches, the Japanese response to the Fukushima Daiichi accident provides valuable lessons for the United States.
The committee did not have the time or resources to perform an in-depth examination of U.S. preparedness for severe nuclear accidents.4 Also, many U.S. agencies are still in the process of developing lessons learned from the Fukushima Daiichi accident. The committee engaged in discus-
Disaster Relief and Emergency Assistance Act, Public Law 93-288, as amended, 42 U.S.C. §§ 5121 et seq.).
4 See Chapter 1, Sidebar 1.2 for a definition of “severe accident.”
sions with several U.S. agencies that have emergency management responsibilities (see Sidebar 6.3) to become better informed about these ongoing efforts: the U.S. Nuclear Regulatory Commission (USNRC), the Federal Emergency Management Agency (FEMA), and the U.S. Environmental Protection Agency (USEPA) (see meeting agendas in Appendix B). In addition, the committee requested information from the U.S. Centers for Disease Control and Prevention (CDC).
This chapter is organized into five sections. The first and second sections aim to put the radiological consequences of the Fukushima Daiichi accident and the difficulties in responding to the accident due to the competing natural disasters—the earthquake and tsunami—into perspective. The third section provides a brief description of the offsite emergency response during the first few days of the Fukushima Daiichi accident. The fourth section discusses some key issues that arose from the committee’s analysis of emergency management in Japan. The fifth and final section provides the committee’s lessons learned for nuclear emergency preparedness in the United States. These lessons learned are presented as findings and recommendations and are directed to the U.S. nuclear industry, state and local governments, and federal agencies with emergency preparedness responsibilities.
The Fukushima Daiichi accident is one of the major accidents in the history of commercial nuclear power. The accident resulted in the most extensive release of radioactive materials into the environment since the 1986 Chernobyl accident in Ukraine. Radioactive releases into the environment started on March 12, 2011, and the significant discharge phase ended at midnight on March 25 (IRSN, 2011). However, minimal releases of radioactive material to the atmosphere continued until December 2011 when cold shutdown of the last impacted reactor at the Fukushima Daiichi plant was achieved (Brumfiel, 2011). Releases to the ocean have continued to the present.
Both the Fukushima Daiichi and Chernobyl nuclear accidents were designated as Category 7 on the International Atomic Energy Agency’s (IAEA’s) International Nuclear and Radiological Event Scale.5 However, the physical health-related radiological consequences of the Fukushima Daiichi accident are less severe than those for the Chernobyl accident for four main reasons:
5 Category 7 is the highest level of the scale.
In the United States, state and local governments have the primary responsibility for making protective action decisions and communicating health and safety instructions to affected populations during a nuclear power plant accident. As laid out in the National Response Framework (NRF; see Sidebar 6.2), a number of federal agencies also play an important role in responding to the accident (USDHS, 2013).
USNRC and FEMA
The USNRC and FEMA are the primary federal agencies responsible for radiological emergency preparedness in the United States. The USNRC is responsible for ensuring that nuclear plants are prepared for radiological emergencies. The USNRC coordinates with FEMA, which oversees state and local agencies’ preparedness for offsite actions. FEMA also provides guidance and support to local and state authorities through its Radiological Emergency Preparedness (REP) program (FEMA, 2013b).
It is not practical for emergency plans to address every possible combination of events (no matter how unlikely) or to present every possible action that can or should be taken in response to an evolving event. Instead, a “planning basis” is available for nuclear power plant events in the United States and is described in a USNRC and USEPA Task Force report (USNRC and USEPA, 1978). The planning basis is utilized in the joint USNRC and FEMA document “Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of Nuclear Power Plants” (USNRC and FEMA, 1980). This document is currently undergoing review; a revised draft is expected to be available for public comment in November 2014.
One of the USEPA’s roles in radiological emergency preparedness is to establish protective action guidelines (PAGs) and provide guidance on implementing them, including recommendations on protective actions. USEPA’s PAGs are expressed in terms of projected doses at which protective actions should be taken to reduce or eliminate exposures (USEPA, 2013). In setting the range of
values for its PAGs, USEPA considered the following four principles (Conklin and Edwards, 2000):
1. Avoid acute radiation health effects.
2. Minimize the risk of delayed health effects.
3. Dose values should not be higher than justified by a cost-benefit analysis.
4. Risks to health from implementing the protective action should not be greater than the risk from the dose avoided.
Emergency responders can use the PAGs for any radiation incident involving relatively significant releases of radioactive materials, including nuclear power plant accidents for the early and intermediary phases.
The CDC’s roles in radiological emergency preparedness include:
1. Providing guidance to state and local governments on the health effects from exposure to radiation and guidance on how to minimize adverse health effects, including psychological health effects from exposure to radiation.
2. Providing medical treatment of exposed individuals and epidemiological surveillance of exposed populations.
3. Participating in the Advisory Team for Environment, Food and Health, a radiological emergency response group tasked with issuing protective action recommendations to prevent or minimize radiation exposure through ingestion by preventing or minimizing contamination of milk, food, and water.
USDOE’s role in a radiological emergency is to coordinate federal environmental radiological monitoring and produce predictive plume models and dose assessments. USDOE makes use of a variety of emergency response assets to estimate the probable or actual spread of radioactivity in the environment. The assets include the National Atmospheric Release Advisory Center (NARAC) for plume and deposition modeling and the Aerial Measuring System (AMS) for measurements of ground deposition with aircraft-mounted detectors. USDOE can create a federal radiological monitoring and assessment center (FRMAC) to help integrate consequence management resources and coordinate the development of a common operating framework.
1. Radioactive releases from the Fukushima Daiichi accident (approximately 100-500 Petabecquerel [PBq]6 of iodine-131 and 6-20 PBq of cesium-1377; UNSCEAR, 2013b) are estimated to be less than 15 percent of those from Chernobyl (approximately 1,760 PBq iodine-131 and 85 PBq cesium-137; UNSCEAR, 2011; Povinec et al., 2013).
2. Prevailing winds at the time of the accident appear to have blown about 80 percent of the radioactive material released from the Fukushima Daiichi plant out to the Pacific Ocean (Kawamura et al., 2011; Morino et al., 2011). The majority of the radioactive material deposited over land was dispersed along a track stretching about 50 km to the northwest of the plant. In contrast, radioactive material from Chernobyl was largely deposited over land (UNSCEAR, 2011).
3. Evacuation of those living in proximity (within 3 km) to the Fukushima Daiichi plant was ordered a few hours after the accident began and at least 12 hours before major releases of radioactive materials from the reactors started (Investigation Committee, 2011). At Chernobyl, evacuations started almost a day after the accident began at which point releases had already started (NEA, 2002; UNSCEAR, 2011).
4. Government restrictions put into place after the Fukushima Daiichi accident kept most contaminated foodstuffs off the market (IAEA, 2011). After the Chernobyl accident there were long delays in implementing appropriate food restrictions in some local areas (UNSCEAR, 2011).
The grave consequences of the Chernobyl accident included the immediate deaths of 28 first responders and firefighters from acute radiation sickness and an epidemic of thyroid cancer in children in Ukraine and neighboring countries.8 With respect to the Fukushima Daiichi accident, there is general agreement in the scientific community that no worker received a dose that resulted in acute radiation death or sickness. Also, doses received by members of the public are estimated to be generally low; therefore, any increase in an individual’s risk of developing cancer in the
6 Becquerel (Bq) is the international (SI) name for the unit of activity; one Bq is equal to one disintegration per second, or 2.7 × 10–11 curies (Ci). 1 PBq = 1.0 × 1015 Bq.
7 The entire inventory of the Fukushima Daiichi Units 1-3 was estimated to be 6,000 PBq iodine-131 and 700 PBq cesium-137 (UNSCEAR, 2013b).
8 About 6,000 excess thyroid cancers were reported up to the year 2005 and many more were projected in the future resulting from exposure to radioactive iodine releases during the Chernobyl accident, mostly through ingestion of contaminated cow’s milk (Cardis et al., 2006; UNSCEAR, 2011).
future is also low (UNSCEAR, 2013b; WHO, 2013; Steinhauser et al., 2014).9
According to reports by the World Health Organization (WHO, 2013) and the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR, 2013b), most people in Fukushima Prefecture received an effective dose10 between 1 and 10 mSv in the first year following the accident. People from Namie, which is located inside the evacuation zone, and Iitate, which is located 40 km (25 miles) northwest of the Fukushima Daiichi plant, may have received the highest effective doses; those doses are estimated to be between 10 to 50 mSv, all delivered in the first year.11 However, infants in Namie were thought to have received higher thyroid radiation doses—between 100 and 200 mSv. The authors of the WHO (2013) report conclude that
The present results suggest that the increases in the incidence of human disease attributable to the additional radiation exposure from the Fukushima Daiichi nuclear power plant accident are likely to remain below detectable levels.
Nevertheless, the government of Japan has launched a 30-year-long health survey of the 2 million residents of Fukushima Prefecture. The survey includes pediatric thyroid monitoring (Yasumura et al., 2012).
This discussion of the physical health-related radiologic consequences
9 This conclusion is based on the linear no-threshold (LNT) model of risk assessment. According to this model, the risk of developing cancer is proportional to dose received, and even a small dose can result in a small increase in lifetime risk of developing cancer. Using risk-projection models, estimates of the number of cancer cases and deaths possibly attributable to the Fukushima Daiichi accident globally or locally have been published in peer-reviewed journals (Ten Hoeve and Jacobson, 2012; Beyea et al., 2013; Evangeliou et al., 2014). These estimates, which should be considered preliminary, are based on LNT risk models developed by the U.S. National Academy of Sciences’ Committee on the Biological Effects of Ionizing Radiation (BEIR VII) (NRC, 2006a). The central estimates range from a few hundred to 1,700 cases depending on the specific LNT model used and do not fully account for uncertainties in the model at low doses. Future revisions to the estimates are likely as doses from the Fukushima Daiichi accident are better assessed (similar to the Chernobyl dose assessments; see Cardis et al., 2006).
10 Effective dose, expressed in millisieverts (mSv), is a dose parameter used to normalize partial-body radiation exposures relative to whole-body exposures to facilitate radiation protection activities (ICRP, 1991). For nuclear power plant accidents where populations are exposed primarily to gamma radiation, such as occurred as a result of the Fukushima Daiichi accident, whole-body dose expressed as effective dose and reported in millisieverts and organ absorbed dose reported in milligrays are numerically equivalent (NRC, 2006a). For consistency throughout this chapter, discussions of dose are in terms of effective dose and reported in millisieverts.
11 For comparison, the average radiation background in Japan is 2.4 mSv/yr. See http://www.jaea.go.jp/04/ztokai/kankyo_e/kaisetsu/expln_1.html. Last accessed on June 12, 2014.
TABLE 6.1 Chronologies of Evacuation and Shelter-in-Place Orders Following the Fukushima Daiichi Accident
|Date in 2011 (time)||Distance from Plant||Ordersa||Area Designation|
|March 11 (20:50)||2 km||Compulsory evacuation issued by the Fukushima prefectural government||Restricted Zone|
|(21:23)||3 km||Compulsory evacuation||Restricted Zone|
|March 12 (05:44)||10 km||Compulsory evacuation||Restricted Zone|
|(18:25)||20 km||Compulsory evacuation||Restricted Zone|
|March 15||20-30 km||Shelter in home||Evacuation Prepared Area|
|March 25||20-30 km||Self-evacuation||Evacuation Prepared Area|
|April 22||Areas with dose >20 mSv/yr||Evacuation within 1 month||Deliberate Evacuation Area|
|June 16||Hotspots with dose >20 mSv/yr||Recommended for evacuation||Specific spots recommended for evacuation|
|September 30||20-30 km||Lifted order to shelter indoors or self-evacuate||Lifting of Evacuation Prepared Area|
a Issued by the central government unless otherwise stated; order unless otherwise stated.
SOURCE: Adapted from Hasegawa (2013).
of the Fukushima Daiichi accident is not intended to downplay other severe long-term health impacts. Of the approximately 150,000 people who were evacuated as a result of the accident12 (UNSCEAR, 2013b), over 80,000 (WNA, 2014) still lived in shelters or temporary locations 3 years after the accident with continuing uncertainties about the future. The difficulties in
12 About 78,000 people living within a 20-km radius of the Fukushima Daiichi plant and 62,000 people living between 20 and 30 km from the plant were evacuated during the first few days of the accident. In April 2011, the government of Japan recommended the evacuation of about 10,000 more people living farther to the northwest of the plant (UNSCEAR, 2013b). See Table 6.1 for the evacuation time line.
evacuees’ daily lives, possible separation from family members, and loss of property and business or employment are further complicated by the fear of developing cancer from accident-related radiation exposures and the societal stigma resulting from those exposures (NRA, 2013b). As with the Chernobyl accident, mental health effects, which include depression, anxiety, and post-traumatic symptoms, are considered to be the largest public health problem from the accident (González et al., 2013; Bromet, 2014).
The environmental and economic consequences of the accident are also severe. About 13,000 km2 of land (about the size of the U.S. state of Connecticut) are contaminated such that the average annual dose to occupants would exceed the 1-mSv/yr long-term cleanup goal (Chen and Tenforde, 2012).13 Cleanup of such a large area is proving to be challenging due to the limited effectiveness of decontamination techniques (Yasutaka et al., 2013), lack of short- or long-term plans for disposal of the radioactive waste created by cleanup, and ongoing negotiations among stakeholders about acceptable radiation dose criteria for resettlement. The final determination of how much residential land will be off limits indefinitely has still not been made.14 Return of evacuated persons—although a high priority of the Japanese government—remains an unresolved issue 3 years after the accident.
Emergency response to the Fukushima Daiichi accident was greatly inhibited by the widespread and severe destruction caused by the Great East Japan Earthquake and tsunami: local electrical power and regional communication infrastructure were knocked out and the transportation infrastructure (roads, bridges, ports, and railroads) was damaged. Japan is known to be well prepared for natural hazards; however, the earthquake and tsunami caused devastation on a scale beyond what was expected and prepared for. More than 20 prefectures were affected by the natural disaster. The National Police Agency of Japan reports 15,883 confirmed deaths and 2,652 people missing due to the earthquake and tsunami. Damage to buildings was extensive: over 126,000 buildings totally collapsed and about 1 million buildings were partially damaged (National Police Agency of Japan, 2014).
Responses to the earthquake and tsunami diverted emergency response
13 The IAEA has recommended a short-term goal of achieving effective doses of 1-20 mSv per year with the ultimate goal of achieving residual effective doses at or below 1 mSv per year (IAEA, 2013b).
14 There are areas where the estimated annual dose level is over 50 mSv per year due to cesium-137 (30-year half-life) and cesium-134 (2-year half-life) deposition. According to IRSN (2012b), the population’s return “seems barely feasible in the long term.”
teams that could have otherwise focused on responding to the Fukushima Daiichi accident. Immediately after the earthquake and tsunami, the government established an emergency response team headed by the prime minister. (The prime minister also acted as the director-general for the offsite response to the nuclear accident.) Within a day of the disaster, the Ministry of Defense ordered the dispatch of the country’s military, the Japanese Self-Defense Forces (SDF), which included 110,000 active and reserve troops, along with 28,000 members of the National Police Force as well as the Fire and Disaster Management Agency (Carafano, 2011). These three forces were also called on during the period March 14-17 to help inject water into the Fukushima Daiichi plant’s cooling systems and spent fuel pools (NERHQ-TEPCO, 2011). In addition, SDF provided air transport within the 20-km evacuation zone to people who needed help to evacuate (Mizushima, 2012). Similarly, the national police assisted with environmental radiation monitoring (NERHQ-TEPCO, 2011), and the Japanese Red Cross Society provided medical and psychological support to earthquake and tsunami victims as well as those affected by the nuclear accident.
In addition to the overwhelming relief demands on the emergency response teams, which had to deal with three simultaneous disasters of unexpected scale, emergency response to the Fukushima Daiichi accident was conducted with limited information on the status of the nuclear plant itself. As described in Chapter 4 of this report, many monitoring and control systems at the plant were not functional because of tsunami-related flooding. Additionally, some offsite instrumentation also was not functional. Consequently, decisions on protective actions for affected offsite populations (e.g., evacuations, sheltering in place, and KI distribution) were made under great stress and great uncertainty about the status of the plant, accident progression prospects, and projected doses to nearby populations.
The following sections describe the offsite emergency response to the accident at the Fukushima Daiichi plant.
6.3.1 Declaration of Emergency
Immediately after the arrival of the second (main) tsunami wave at 15:36-15:37 on March 11 (see Sidebar 3.1 in Chapter 3), TEPCO, in accordance with Article 10, Paragraph 1, of the Act on Special Measures Concerning Nuclear Emergency Preparedness (Cabinet Secretariat, Government of Japan, 1999), informed the Nuclear Industry Safety Agency (NISA) of the plant’s total loss of alternating current (AC) power. This notification
was made at 15:42 on March 11. There are two different accounts of the step that followed:
• By one account (Investigation Committee, 2011), NISA, in consultation with the Ministry of Economy, Trade and Industry (METI), determined at 16:36 on March 11 that the incident rose to the level of a nuclear emergency situation as defined in Article 15, Paragraph 1, of the Act. This Act calls for the Japanese prime minister to immediately give public notice of the occurrence of a nuclear emergency situation.
• By another account, TEPCO informed NISA at 16:45 on March 11 that the situation required the Article 15 public notice.
In either case, at around 17:42 on March 11, NISA and METI reported the situation to the prime minister and provided him with a draft public notice. The prime minister gave the required public notice at 19:03 on March 11.
Authorities in Japan acted immediately to reduce the consequences of potential releases of radioactive materials from the Fukushima Daiichi plant. Their actions were to be coordinated through the Nuclear Emergency Response Headquarters (NERHQ), which was established near the prime minister’s office in Tokyo and was led by the prime minister. In addition, the local NERHQ was established in Fukushima Prefecture about 5 km west of the Fukushima Daiichi plant and was led by METI’s senior vice-minister. However, full operation of the local NERHQ was delayed until about March 15 (JNES, 2013). This delay was due to the lack of electrical power and damage to highways and roads, which made local travel difficult.15 Because of this delay, coordination between the national and local governments for ordering, implementing, and confirming evacuations and other protective actions was difficult.
6.3.2 Issuance of Protective Actions
Instrumentation that would normally have been used to inform protective-action decisions following the accident were unavailable due to the loss of electrical power and damage from the earthquake and tsunami. This instrumentation included
• Twenty-four radiation monitoring stations on the Fukushima
15 The alternative location for the offsite center (in the Minamisoma City Hall) was already being used as an emergency response center for the earthquake and tsunami. The local NERHQ was therefore established in the Fukushima Prefectural Building.
Daiichi plant site; 23 of these stations were rendered nonfunctional by the tsunami (WNA, 2014).
• The Emergency Response Support System (ERSS), which provides data on plant status to multiple offsite centers. This system malfunctioned immediately after the accident (NERHQ-TEPCO, 2011). Consequently, critical information about the status of the Fukushima Daiichi plant could not be obtained.
• The System for Prediction of Environmental Emergency Dose Information (SPEEDI) is used during emergencies to predict atmospheric concentrations of radioactive materials, dose rates, and environmental exposures. These predictions are used to inform decisions by authorities on protective actions.16 The ERSS feeds information on radioactive release sources to SPEEDI; but, as noted previously, ERSS was not functional.
Reliable real-time estimates of sources and magnitudes of radioactive material releases from the Fukushima Daiichi plant were therefore unavailable.
As discussed in Chapter 4, some releases of radioactivity to the atmosphere from the plant occurred through uncontrolled pathways (see also Narabayashi et al., 2012). An instrument at the main gate of the plant produced a continuous record of gamma dose rate from these releases (NERHQ-TEPCO, 2011) and cars with measuring instruments produced some scattered measurements elsewhere on the site. However, these data could not be analyzed in real time.
The Ministry of Education, Culture, Sports, Science and Technology (MEXT) obtained some measurements on March 15 from a car located 20 km to the northwest of the plant (NERHQ-TEPCO, 2011). A number of monitoring instruments were also set up beyond the 20-km evacuation radius starting on March 16 (NERHQ-TEPCO, 2011). Gross gamma dose-rate measurements from these instruments were adequate to determine whether populations should be moved from already-contaminated areas.
Airborne measurements of ground contamination levels were made by the U.S. Department of Energy (USDOE) starting March 17 (Lyons and
16 Weather forecasting is uncertain, so any projection of plume transport using SPEEDI becomes increasingly uncertain as the forecast time for the projection increases. Also, the timing of multiple, prolonged releases with respect to wind patterns complicates predictions. As a result, projections with SPEEDI or other similar systems can only be probabilistic. The uncertainties increase in situations with multiple releases occurring at apparently random times.
Given the sparse information on the status of the plant and uncertainties about projected doses, decision makers who issued protective actions showed a preference for evacuation of populations located near the plant rather than sheltering in place.
220.127.116.11 Evacuation Orders
Several evacuation orders were issued following the prime minister’s declaration of a nuclear emergency (see Table 6.1). The evacuation zones were gradually expanded over time, and residents were ordered to evacuate repeatedly from one place to another. Prior to instructions from the NERHQ (see Section 6.3.1), the governor of Fukushima Prefecture instructed Okuma Town and Futaba Town—the two towns nearest to the Fukushima Daiichi plant (see Figure 6.1)—to evacuate residents living within a 2-km radius of the Fukushima Daiichi plant. Approximately 30 minutes later, the NERHQ instructed the Fukushima Prefectural governor and all relevant local governments to issue an evacuation order to citizens within a 3-km radius of the plant and to issue a shelter-in-place order to citizens between 3 and 10 km of the plant. These evacuation orders were preemptive; there were no data at the time indicating that there had been a release of radioactive material from the plant or that such a release was imminent.
Following instructions by the prime minister to the heads of relevant municipalities, the evacuation area was increased to a 10-km radius the morning of March 12 because of fears that potentially large quantities of radioactive materials would be released. The evacuation zone was further increased to 20 km that afternoon following the hydrogen explosion at Unit 1 (see Sidebar 3.1 in Chapter 3). Fukushima Prefecture, Okuma town, Futaba town, Tomioka town, Namie town, Kawauchi town, Naraha town, Minamisoma city, Tamura city, and Katsurao village were among the municipalities evacuated (NERHQ-TEPCO, 2011). An estimated 78,000 people evacuated from the 20-km-radius zone around the plant
17 Measurements were made from altitude bands of 152-305 m (helicopter) and 550-700 m (fixed-wing aircraft). Each aircraft used detectors equipped with a total of 12 large-volume (5 cm × 10 cm × 40 cm) sodium iodide scintillator crystals.
18 Such measurements are made to identify areas that should be subject to long-term evacuations because of contamination by the long-half-life isotopes cesium-134 (2-year half-life) and cesium-137 (30-year half-life). These measurements are not intended to inform decisions on short-term precautionary evacuations (i.e., evacuations to protect populations from exposures to high-dose-rate, short-lived fission products, or decisions to advise populations to take KI to prevent thyroid uptake of inhaled radioactive iodine when evacuations cannot occur in time to avoid such inhalation).
FIGURE 6.1 Evacuation zones established by the Japanese government following the Fukushima Daiichi accident. Fukushima Nuclear Power Plant (No. 1) is the Fukushima Daiichi plant; Fukushima Nuclear Power Plant (No. 2) is the Fukushima Daini plant. SOURCE: Adapted from Hasegawa (2013).
(UNSCEAR, 2013b). This area was designated as a “Restricted Zone” with entry initially prohibited.19
The hydrogen explosions in Unit 1 (15:36 on March 12), Unit 3 (11:01 on March 14), and Unit 4 (06:14 on March 15) (see Chapter 3, Sidebar 3.1) led the prime minister to issue new instructions to the heads of relevant local governments on March 15 including Fukushima Prefecture, Okuma town, Futaba town, Tomioka town, Namie town, Kawauchi town, Minamisoma city, Katsurao village, Hirono town, and Iitate village (see Figure 6.1), to order residents within the 20- to 30-km radius from the plant to shelter in place in what was designated as an “Evacuation Prepared Area.”
19 Some progress has been made with respect to the resettlement of parts of this area (METI, 2013).
Approximately 60,000 people lived within the 20- to 30-km shelter-in-place zone (UNSCEAR, 2013b). On March 25, these residents were advised by the government to begin voluntary evacuations.
On April 22, 2011, the central government issued a new evacuation order to residents of Iitate, located outside the 20-km radius evacuation zone, where high radiation levels had been detected. Residents of that village were given 1 month to evacuate. The area was designated as a “Deliberate Evacuation Area.”
At this point onward the government switched from communicating evacuation orders on the basis of distance from the plant to using a threshold radiation dose of 20 mSv/yr as a basis for evacuation (Hasegawa, 2013). In June 2011, the government began to identify hotspots20 where radiation levels exceeded this 20-mSv/yr threshold. These hotspots were named “Specific Spots Recommended21 for Evacuation.” These hotspots were more than 20 km away from the Fukushima Daiichi plant and outside the Deliberate Evacuation Area (UNSCEAR, 2013b).
18.104.22.168 Potassium Iodide Distribution
In addition to evacuation and shelter-in-place orders, residents leaving the 20-km Restricted Zone were instructed to take potassium iodide (KI). This instruction was issued on March 16, four days after major releases of radioactive iodine (iodine-131) had begun and after about half of the iodine release had occurred (TEPCO, 2012b, Fig. 27). This was also 4 days after residents within the Restricted Zone were instructed to evacuate and a day after residents in the 20- to 30-km Evacuation Prepared Area were instructed to shelter in place. Upon issuing this instruction, KI was made available for distribution. The KI consisted of 1.51 million pills for 750,000 people and 6.1 kg powder for 120,000–180,000 people. However, the KI was likely not distributed because the evacuation had already been completed (Hamada et al., 2012).
On March 15, four towns close to the plant, Futaba, Tomioka, Iwaki,22 and Miharu, distributed in-stock KI pills to local residents without awaiting distribution instructions from the government. Futaba and Tomioka also instructed their residents to take the pills (Hayashi, 2011).
20 Hot spots are defined based on radioactive contamination levels. They are regions where contamination levels significantly exceed those in surrounding areas.
21 In other words, evacuations in these areas were not ordered.
22 Iwaki is located south of the area shown in Figure 6.1.
22.214.171.124 Food Interdictions
On March 15, 2011, high levels of radioactive iodine (iodine-131) and radioactive cesium (cesium-134, -137) were detected in topsoil and vegetation near the Fukushima Daiichi plant (Hamada et al., 2012).23 The Nuclear Safety Commission (NSC) advised that monitoring surveys of food and water should begin immediately. Food and water samples were collected beginning on March 16, 2011. On March 17, the Ministry of Health, Labor and Welfare (MHLW) set regulatory limits for contaminated food and water; these limits were stipulated as provisional regulatory values (PRVs).
PRVs were adopted from the index values preset by NSC except for radioactive iodine (iodine-131) in water and milk ingested by infants and in seafood24 (Hamada and Ogino, 2012). PRVs for foodstuffs and drinkable liquids contaminated with radioactive cesium (cesium-134, -137), uranium, plutonium, and some other transuranic isotopes were based on an effective dose limit not to exceed 5 mSv/yr (Hamada and Ogino, 2012). The Food Safety Commission of Japan decided on March 20, 2011, that these PRVs were effective enough to ensure public safety. These PRVs were applied in various districts of Fukushima, Ibaraki, Chiba, Miyagi, Tochigi, Iwate, Gunma, and Kanagawa Prefectures starting on March 21 (FSC, 2011; Hamada and Ogino, 2012; IRSN, 2012b).
These initial PRVs were in place until March 31, 2012. New regulatory values went into effect on April 1, 2012. These new values were expressed as radioactive concentrations of cesium-134 and -137, but also considering the contributions of strontium-90, plutonium-238, -239, -241, and ruthenium-106, not to exceed a committed effective dose of 1 mSv/yr (Hamada and Ogino, 2012).
6.3.3 Accident Recovery
From March 11, 2011, to August 2011, implementation of an integrated recovery plan was hampered by administrative delays. In particular, time was needed to establish the required administrative structures, regulations, and a budget framework for those recovery actions that were not covered in existing disaster management plans (Hardie and McKinley, 2013). Decontamination activities during this period were conducted outside of the evacuation zones with a focus on high-sensitivity areas, such as schools
23 Hamada et al. (2012) do not specify the location where high levels of cesium were found.
24 Contamination of foodstuffs and liquids with iodine-131 became less of a public health concern with time owing to that isotope’s short half-life (approximately 8 days). This was not the case for foodstuffs and liquids contaminated with cesium-134 and cesium-137 which have much longer half-lives.
In August 2011, the Japanese government passed the Act on Special Measures Concerning the Handling of Radioactive Pollution.25 Pursuant to the Act, Japanese agencies developed a framework and guidance for remediating contaminated areas. These guidelines cover methods for surveying and measuring contamination levels as well as strategies for decontamination and storage of contaminated materials (Yasutaka et al., 2013). The Act took full effect in January 2012; it established JAEA as the responsible organization for coordinating the development of a technical basis for the regional remediation plan to be developed under the Act (Hardie and McKinley, 2013).
According to the Act, contaminated areas were to be grouped into two categories:
• Special decontamination areas. These areas comprise the Restricted Zone (i.e., areas within 20 km of the plant) as well as the Deliberate Evacuation Area (i.e., area beyond 20 km where the annual effective dose for individuals was anticipated to exceed 20 mSv26). The national government is responsible for the decontamination of these areas with a goal to reduce annual cumulative doses to less than 20 mSv. The long-term goal is to reduce annual cumulative dose to less than 1 mSv.
• Intensive contamination survey areas. These include all other contaminated areas in which the cumulative radiation dose for individuals was anticipated to range between 1 mSv and 20 mSv annually. Decontamination is to be overseen primarily by local municipalities with the goal to reduce the “air dose rate” to less than 1 mSv/yr.
26 According to ICRP (2011), 1-20 mSv/yr is the reference dose recommendation for exposure situations involving, for example, people living in long-term contaminated areas after a nuclear accident or a nuclear emergency.
FIGURE 6.2 METI projections for land decontamination end states in regions affected by the Fukushima Daiichi accident. Area 1 (green): estimated annual dose level is below 20 mSv; Area 2 (orange): estimated annual dose level is 20-50 mSv; Area 3 (pink): estimated annual dose level is over 50 mSv and residents are not allowed entry. SOURCE: METI. Available at http://www.meti.go.jp/english/earthquake/nuclear/roadmap/pdf/140401MapOfAreas.pdf.
• Area 1: Estimated annual dose level is below 20 mSv and residents can return home temporarily. Evacuation orders within this area are “ready to be lifted.”28
• Area 2: Estimated annual dose level is 20-50 mSv; residents are allowed entry for specific purposes but are ordered to remain evacuated.
• Area 3: Estimated annual dose level is over 50 mSv and residents are legally required to remain outside these areas. Levels are not expected to drop below 20 mSv/yr before about March 2016, five years after the Fukushima Daiichi accident.
Decontamination of these areas involves the cleaning of structures and removal of contaminated soil. The removed soil and other contaminated wastes are being stored at remediation locations or at temporary sites.29 Incineration is being used for volume reduction of some contaminated materials (while meeting applicable emission standards for limiting public exposures) (IAEA, 2014b). Contaminated soil and waste are to be gathered and placed into interim storage facilities until transferred to a long-term disposal site outside of the Fukushima area. The national government aims to have these interim storage facilities in operation by early 2015.30
FINDING 6.1: The Fukushima Daiichi accident revealed vulnerabilities in Japan’s offsite emergency management. The competing demands of the earthquake and tsunami diminished the available response capacity for the accident. Implementation of existing nuclear emergency plans was overwhelmed by the extreme natural events that affected large regions, producing widespread disruption of communications, electrical power, and other critical infrastructure over an extended period of time. Additionally:
28 METI’s April 2014 map of Area 1 has remained unchanged, with few exceptions, since the previous update provided by the agency in August 2013. Therefore METI’s designation of Area 1 as “evacuation orders are ready to be lifted” may be misleading. (See METI maps http://www.meti.go.jp/english/earthquake/nuclear/roadmap/pdf/20130807_01.pdf and http://www.meti.go.jp/english/earthquake/nuclear/roadmap/pdf/140401MapOfAreas.pdf for a direct comparison of the areas.)
30 See https://www.reconstruction.go.jp/english/topics/2013/03/decontamination-process.html. Accessed on July 17, 2014.
• Emergency management plans in Japan at the time of the Fukushima Daiichi accident were inadequate to deal with the magnitude of the accident, requiring emergency responders to improvise.
• Decision-making processes by government and industry officials were challenged by the lack of reliable, real-time information on the status of the plant, offsite releases, accident progression, and projected doses to nearby populations.
• Coordination among the central and local governments was hampered by limited and poor communications.
• Protective actions were improvised and uncoordinated, particularly when evacuating vulnerable populations (e.g., the elderly and sick) and providing potassium iodide.
• Different and revised radiation standards and changes in decontamination criteria and policies added to the public’s confusion and distrust of the Japanese government.
• Cleanup of contaminated areas and possible resettlement of populations are ongoing efforts 3 years after the accident with uncertain completion time lines and outcomes.
• Failure to prepare and implement an effective strategy for communication during the emergency contributed to the erosion of trust among the public for Japan’s government, regulatory agencies, and the nuclear industry.
6.4.1 Lack of Planning for a Severe Nuclear Accident
According to an independent Diet investigation of the Fukushima Daiichi accident (NAIIC, 2012), Japan was not prepared for the severe demands of the triple disaster that occurred on March 11, 2011. Moreover, Japan’s preparedness for a nuclear disaster would have been deficient even if it had occurred in isolation of the natural disasters. The possibility of a reactor-core-damaging event at a nuclear plant in Japan was considered implausible (see Chapter 4). Consequently, planning for such an event was not treated seriously, leaving Japan unprepared for the scope and extent of the required emergency response. For example, the 2007 NSC guide for emergency preparedness describes the basis for establishing an Emergency Planning Zone31 as being a result of a hypothesized situation of releases that could not possibly occur (NSC, 2013). The guide further states that actions such as sheltering in place or evacuation would not be needed outside of the 8- to 10-km radius (NSC, 2013).
The belief that design, engineering, and administrative controls related to nuclear plant operation excluded the possibility of a severe accident may
31 This zone extends out to approximately 10 km from a nuclear plant.
have contributed to the many difficulties faced during implementation of the Japanese government’s Basic Plan for Emergency Preparedness. In addition, the plan did not address contingencies, such as the loss of electrical power and communications, or diversion of response staff (e.g., local police and fire response resources) by competing events such as an earthquake and tsunami (NSC, 2013).
In spite of these limitations, the Japanese government was able to substantially decrease radiation exposure risks to the public using standard protective actions: evacuation, sheltering in place, and food and water interdictions (NSC, 2013). While some KI was distributed by Japanese prefectures and towns near the Fukushima Daiichi plant (see Section 126.96.36.199), it is not clear whether this KI was taken, and, if it was, whether its administration resulted in dose savings.
6.4.2 Lack of Reliable Information to Make Informed Decisions
As noted in Section 6.3.2, SPEEDI was not functioning at its full capacity after the accident.32 Moreover, much of the radioactive material releases from the Fukushima Daiichi plant were through uncontrolled pathways. The loss of onsite power made measurements of radioactive material releases through controlled-release pathways (e.g., through the plant stack) impossible. SPEEDI could still be used to estimate the offsite atmospheric dispersion of radioactive noble gases and iodine using reference release rates (Investigation Committee, 2011). However, these estimates were not always communicated to relevant organizations (e.g., MEXT, NSC, and Fukushima Prefecture) (Investigation Committee, 2011). Also, SPEEDI results were not initially made public; therefore, local governments could not use those results to plan evacuations.33
6.4.3 Uncoordinated Issuance of Protective Actions
As noted in Section 188.8.131.52, a series of evacuation decisions were made as the accident at the Fukushima Daiichi plant unfolded. However, because of the lack of reliable and timely information about radioactive material releases from the plant, coupled with the loss of electrical power and gen-
32 Normally, if the releases were via the reactor stacks and power was available, the rates of release of radioactive materials could be monitored during an accident. See USNRC (2006).
33 SPEEDI estimates were not made public for several reasons. First, officials did not trust the SPEEDI estimates because they were made using assumed scenarios. Second, until March 16, it was not clear whether MEXT or NSC had the responsibility for operating SPEEDI. On March 16, NSC became the responsible organization for operating and maintaining SPEEDI and making its estimates public. The first SPEEDI estimates were made public on March 23 (Investigation Committee, 2011).
eral disruption of infrastructure, these decisions had to be made on an ad hoc basis. The evacuations were considered a precautionary response in light of the uncertainties about the status of the Fukushima Daiichi reactors and the potential offsite doses to surrounding populations if the reactors could not be stabilized.
The protective actions issued during the accident were generally successful due to the combination of good execution by the organizations involved, improvisation, and good luck. With respect to “good luck,” as noted previously, about 80 percent of the radioactivity released from the Fukushima Daiichi plant was transported to the Pacific Ocean (Kawamura et al., 2011; Morino et al., 2011). Nevertheless, according to UNSCEAR (2013b), evacuations reduced by up to a factor of 10 the doses that would have been received by those living in the evacuated areas.
Independent of how successful the evacuation orders were in reducing radiation doses, evacuation instructions issued by the central and local governments lacked coordination, primarily because of disruptions to the communication infrastructure. As a result, many (and perhaps most) residents ordered to evacuate had to do so repeatedly, moving from one place to another (Hasegawa, 2013; Kurihara, 2013).
Additional confusion was caused by a March 16 recommendation issued by the U.S. Department of State that Americans located within about 80 km (50 miles) of the Fukushima Daiichi plant evacuate because of concerns that the situation could worsen (USNRC, 2011d; U.S. Embassy, 2011). This order sowed confusion and anxiety among the Japanese people who were living in this zone, because they had not been told by the Japanese government to evacuate. It also prompts the question “What is the appropriate role of foreign authorities in providing recommendations to its traveling or relocated citizens in a nuclear emergency?” This question is particularly relevant when recommendations are contradictory to those made by the host country government (González et al., 2013). This complex issue needs to be studied and resolved, not only for potential future nuclear accidents or events involving different national governments,34 but also for those involving national and local authorities of the same country.
A major issue with the evacuations during the Fukushima Daiichi acci-
34 Understanding this complex issue is particularly important for future national or international U.S. responses that involve a neighboring country, for example, Canada. The committee was told that the United States and Canada recognize the need for cooperation in developing responses to nuclear plant accidents and other radiological emergencies (FEMA, oral communication with the committee, May 17, 2013).
dent was the lack of detailed planning for vulnerable populations such as the elderly and the hospitalized (Nomura et al., 2013). Tanigawa et al. (2012) describe the chaotic evacuation of bedridden patients, some of whom died before reaching admitting facilities because of evacuation-related trauma or their own medical conditions. Patients could not take personal belongings because of space restrictions, and many patients were transferred several times to different locations over the period of a few months. Some institutions denied entrance to evacuees due to fears that they could contaminate others with radioactivity (Tanigawa et al., 2012; Tominaga et al., 2014). According to a recent article (Tominaga et al., 2014), there was insufficient education and training of emergency responders and physicians on radiation and its health effects.
Healthy elderly were also at risk. Yasumura et al. (2013) analyzed monthly mortality data among the 1,770 institutionalized elderly who were relocated from nursing homes, geriatric health service facilities, and other facilities within the 20-km evacuation zone. They estimated that there were 109 excess deaths among persons in this group.35 The most common cause of death was pneumonia (41 percent), possibly due to low temperatures and poor nutrition (Yasumura et al., 2013). This high number of excess deaths emphasizes the vulnerability of geriatric populations to relocation.
184.108.40.206 Potassium Iodide
As noted in Section 220.127.116.11, little KI was administered to populations living near the Fukushima Daiichi plant because they had already evacuated. Reports indicate that radiation thresholds for KI administration were not exceeded as a result of timely evacuations (UNSCEAR, 2013b). Consequently, discussions of KI’s medical effectiveness during the Fukushima Daiichi accident are irrelevant. However, the efforts to distribute KI during the accident highlight problems with planning for KI distribution and communication and coordination between the national and local governments. These problems were due in part to the loss of the communication infrastructure because of the earthquake and tsunami.
For example, on the night of March 14, the NERHQ and NSC were informed that not all hospitalized patients within the 20-km-radius Deliberate Evacuation Area had actually evacuated. The next day, NERHQ sent a fax to its local NERHQ office advising that these patients should take KI as they evacuate. However, the fax arrived as the local NERHQ office was relocating to the Fukushima Prefectural Office building; it was not discovered until later that evening. The local NERHQ considered it highly
35 The focus here was on those affected by the nuclear emergency, not the earthquake and tsunami.
likely that many other elderly citizens and hospital staff still remained in this zone, so it created an instruction draft advising everyone that remained to take KI. The Fukushima prefectural government had confirmed, however, that there were no remaining elderly citizens or hospital staff within this zone (Investigation Committee, 2011).
6.4.4 Revisions to Radiation Standards
Confusion among residents living in the evacuation zones was exacerbated by the changing radiation safety standards established by the central government in the days and weeks following the accident. As noted in Section 18.104.22.168, from April 22, 2011, onward, the government started communicating relocation and sheltering orders to the public based on a threshold radiation dose of 20 mSv/yr (Hasegawa, 2013). This dose is 20 times higher than the allowable dose limit for public exposures resulting from normal nuclear power plant operations (1 mSv/yr in Japan). Although within the range set by the International Commission on Radiological Protection (ICRP) of 1-20 mSv/yr for “emergency or existing radiation exposure conditions” including nuclear accidents (ICRP, 2007), it was difficult for the public to understand “why the dose limit of 1 mSv/yr, which was valid before the accident, could be exceeded after the accident—at a time when people expect[ed] to be better protected” (González, 2012).
The public’s trust in the government was also affected by the resignation of an academic advisor to the government over the 20-mSv/yr threshold. The advisor judged that this threshold was not acceptable for children (Kai, 2012).36 This example suggests an important lesson: individuals responsible for informing the establishment of radiation protection guidelines during an accident need to agree on the technical criteria for establishing such guidelines in advance, and they also need to support the guidelines subsequently. However, agreeing on criteria that will cover all conceivable situations ahead of time is challenging because of the accepted standard practice of reducing doses to a level as low as deemed reasonably achievable (ALARA). It is not possible to know ahead of time what “reasonably achievable” may mean in a particular situation. Moreover, decisions about “what is reasonable” are themselves somewhat subjective, which can breed suspicion in situations where there is a lack of trust. Nevertheless, the solution is not to defer the establishment of standards or develop comprehensible explanations until a crisis forces action.
Additional public confusion and reported loss of trust in the Japanese government in its management of the response to the accident (Hasegawa,
36 It is understood that for a given radiation dose, children are generally at a higher risk of developing cancer compared with adults (UNSCEAR, 2013a.)
2013) relates to the change in decontamination criteria for evacuees before and during the accident. Prior to March 11, 2011, a decontamination criterion for evacuees was established at 40 Bq/cm2. Using various assumptions, and for practical purposes, this was translated to correspond to a reading of 13,000 counts per minute on a widely used Geiger-Muller radiation survey meter (Ogino and Hattori, 2013). On March 14, 2011, the Fukushima prefectural government decided that only partial decontamination would be undertaken for evacuees with readings between 13,000 and 100,000 counts per minute, and also that full-body decontamination would be undertaken only for evacuees that exceeded 100,000 counts per minute. This revision was made out of two concerns: (1) that too many persons would be considered “contaminated” at low levels (Investigation Committee, 2012)37; and (2) that there were not enough decontamination tools such as tents and water to process the large numbers of persons who would need to be decontaminated (Ogino and Hattori, 2013).38
Another example of changing radiation standards following the accident relates to food interdictions (see Section 22.214.171.124). As the accident evolved, the Japanese government issued PRVs to limit the intake of contaminated food and water. These “provisional” limits aimed to maintain doses to below 5 mSv/yr. These were later revised downward to a maximum permissible dose of 1 mSv/yr (MHLW, 2012, Slide 3).
The bases and reasons for revising these radiation standards continue to be an issue during the recovery phase of the accident. Cleanup of the large areas contaminated by releases of radioactive materials from the Fukushima Daiichi plant is proving to be challenging due to the limited effectiveness of decontamination techniques (Yasutaka et al., 2013) and lack of short- and long-term plans for disposal of the radioactive waste created during the cleanup.
6.4.5 Lack of an Effective Communication Strategy
A wide number of sources indicate that public trust was challenged due to the perception of partial, incorrect, delayed, and ambiguous information about the accident (Nakamura and Kikuchi, 2011; NERHQ-TEPCO, 2011;
37 The written materials reviewed by the committee did not specify the area over which the 13,000 (or 100,000) counts per minute should be measured. However, it may be reasonably presumed that the counts correspond to measurement of an area equal to the active area of the probe used.
38 According to the IAEA, it is unlikely that skin or clothing contamination from radioactive materials released during a severe nuclear accident would pose a significant health concern to offsite populations. Members of the general public can protect themselves from radioactive material on the skin and clothing by taking precautions such as showering and changing their clothes at the first opportunity (IAEA, 2003).
Fitzgerald et al., 2012; Kai, 2012; Ng and Lean, 2012; Figueroa, 2013; Hosono et al., 2013; Tateno and Yokoyama, 2013). According to these sources, the government
• Failed to characterize accurately the conditions at the plant in terms of safe shutdown of reactors and radioactive material releases (Imtihani and Mariko, 2013);
• Rejected as premature a report by a press secretary, later admitted to be correct, that there had been a reactor meltdown early after the tsunami hit the plant (Nakamura and Kikuchi, 2011);
• Withheld for considerable time rough (albeit questionable) predictions of doses based on hypothetical release magnitudes (Tateno and Yokoyama, 2013);
• Provided vague explanations about associated risks (RJIF, 2014); and
• Did not provide adequate information on food contamination and internal exposure (Tateno and Yokoyama, 2013).
Not surprisingly, these perceptions of mistrust increased the tendency of the media to seek out alternative, nongovernmental information sources. The messages from these alternative sources sometimes conflicted with, or appeared to be in conflict with, government statements (Sasakawa, 2012). This was the first major nuclear power plant accident in the Internet age. According to one view, Internet information about the risks of radiation exposure increased public concerns about the risks to children (Kai, 2012). However, a survey of residents living within 300 km of the Fukushima Daiichi plant indicates that the public was skeptical about Internet information and placed NHK39 at the top of the list of credible sources (Tateno and Yokoyama, 2013).
FINDING 6.2: The committee did not have the time or resources to perform an in-depth examination of U.S. preparedness for severe nuclear accidents. Nevertheless, the accident raises the question of whether a severe nuclear accident such as occurred at the Fukushima Daiichi plant would challenge U.S. emergency response capabilities because of its severity, duration, and association with a regional-scale
39 NHK is Japan’s national public broadcasting organization known in English as Japan Broadcasting Corporation.
natural disaster. The natural disaster damaged critical infrastructure and diverted emergency response resources.
RECOMMENDATION 6.2A: The nuclear industry and organizations with emergency management responsibilities in the United States should assess their preparedness for severe nuclear accidents associated with offsite regional-scale disasters. Emergency response plans, including plans for communicating with affected populations, should be revised or supplemented as necessary to ensure that there are scalable and effective strategies, well-trained personnel, and adequate resources for responding to long-duration accident and/or disaster scenarios involving
• Widespread loss of offsite electrical power and severe damage to other critical offsite infrastructure, for example, communications, transportation, and emergency response infrastructure;
• Lack of real-time information about conditions at nuclear plants, particularly with respect to releases of radioactive material from reactors and/or spent fuel pools; and
• Dispersion of radioactive materials beyond the 10-mile emergency planning zones for nuclear plants that could result in doses exceeding one or more of the protective action guidelines.
RECOMMENDATION 6.2B: The nuclear industry and organizations with emergency management responsibilities in the United States should assess the balance of protective actions (e.g., sheltering in place, evacuation, relocation, and distribution of potassium iodide) for offsite populations affected by severe nuclear accidents and revise the guidelines as appropriate. Particular attention should be given to the following issues:
• Protective actions for special populations (children, ill, elderly) and their caregivers;
• Long-term impacts of sheltering in place, evacuation, and/or relocation, including social, psychological, and economic impacts; and
• Decision making for resettlement of evacuated populations in areas contaminated by radioactive material releases from nuclear plant accidents.
The Fukushima Daiichi accident revealed that existing Japanese emergency response plans for dealing with nuclear accidents were inadequate, and it brought to the surface problems with the coordination and decisionmaking processes used by government and industry officials. The difficul-
ties in responding to the accident were exacerbated by the lack of reliable, real-time information on the status of the plant, accident progression, and projected doses to nearby populations.
The committee did not have the time or resources to perform an in-depth examination of U.S. preparedness for severe nuclear accidents. However, the accident raises the question of whether emergency preparedness in the United States would be challenged if a similar-scale nuclear accident were to happen domestically when emergency responses were diverted to deal with concurrent disasters.
6.5.1 Emergency Response Planning Around Nuclear Power Plants
Because of the severe damage to the reactors at the Fukushima Daiichi plant, actions that would normally be associated with the “early phase” of a nuclear incident extended over many days to weeks, rather than the expected hours to days that inform nuclear emergency planning in the United States (see Sidebar 6.1 for definitions of accident phases). Consequently, response staff and resources were needed for an extended period of time. Furthermore, the earthquake and tsunami consumed local police and fire response resources which, similar to the United States, play an integral role in conducting orderly evacuations and providing other emergency services.40 Medical and evacuation center staff and resources were overextended by the demands of responding to two natural disasters (the earthquake and tsunami) and the lengthy time frame of the unfolding accident.
The committee has identified three challenges from the Fukushima Daiichi accident that have the potential to compromise offsite emergency responses to Fukushima-scale events in the United States. These challenges are described below:
1. Widespread loss of offsite electrical power and severe damage to critical infrastructure: The impact to the communications, transportation, and electrical power infrastructure caused by the earthquake and tsunami affected the ability of the Japanese national government to communicate with prefectural and local governments. These types of impacts might also be seen in a response to a Fukushima-scale event in the United States. It is worth examining contingency planning for events that include major disruptions to communications, transportation, and the electrical power infrastructure and for which state, local, medical, and emergency reception center staff and resources are diverted by a competing disaster.
40 In the United States, for example, many plans call for state or local first responders to conduct contamination screening during an event to support the movement of evacuees to an emergency reception center.
2. Lack of real-time information about conditions at the Fukushima Daiichi plant: Questions related to the reliability and ease of use of information from radiation monitoring equipment in Japan proved to be largely irrelevant during the accident response because onsite and offsite instrumentation was not functional, at least at full capacity.41 As a result, information was too sparse for creation of timely and reliable analyses of the plant status and future prospects. Many emergency response decisions were made without a firm basis of situational knowledge. This led to a preference for evacuation versus sheltering in place for populations affected by the accident. It is worth examining whether, given the same set of circumstances in the United States, the information needed to select protective actions would be available—or whether failures, for example, of the Safety Parameter Display System (SPDS),42 Emergency Response Data System (ERDS),43 or normal telephone system would result in a similar loss of information. The USNRC is aware of this issue and believes that its ERDS modernization program, which was under way before the Fukushima Daiichi accident, is sufficient (USNRC, 2011c).44 The Committee did not review this modernization program.
3. Dispersion of radioactive material beyond the 10-mile EPZ resulting in doses exceeding one or more protective action guidelines (PAGs): Radiation doses exceeded some PAGs beyond 30 km (18.6 miles) from the Fukushima Daiichi plant. This suggests that, if a similar-scale accident were to occur in the United States, the 10-mile plume exposure pathway emergency planning zone (EPZ) currently established by the USNRC45 may prove inadequate.46 Given the Japanese experience, it would be worthwhile for the USNRC, FEMA, state and local entities, and industry to review and assess the scalability and effectiveness of emergency response plans for events that lead to significant potential radiation doses and radioactive contamination extending beyond the 10-mile EPZ. This would include reviewing the plans for issuing protective actions such as evacuation, sheltering in
41 See Appendix M for the committee’s opinions on how access to timely and reliable information to support decision making can be improved.
42 The SPDS displays a set of plant parameters from which the plant operators can assess the safety status of plant operation (USNRC, 1980b).
43 The ERDS provides electronic transmission capability of a limited set of parameters from the plant computer to the USNRC during an emergency.
44 The aim of the USNRC’s ERDS modernization program is to ensure that the agency can receive data from all affected reactor units during a multiunit event.
45 Emergency response plans are in place for the plume exposure pathway EPZ to avoid or reduce dose from potential radiation exposure from the release of radioactive materials after a nuclear event. These plans provide the structure for implementing protective actions such as evacuation, sheltering in place, and the use of KI.
46 As noted in Section 126.96.36.199, the U.S. government advised Americans within 50 miles of the Fukushima Daiichi plant to evacuate; similarly, a domestic accident could require evacuation and other protective actions in populations at a similar distance from a U.S. nuclear plant.
place, and use of KI47 to populations that reside beyond the 10-mile EPZ and testing the effectiveness and scalability of these plans by performing regular exercises.
There is a need to ensure that emergency response plans in the United States include scalable and effective strategies, well-trained personnel, and adequate resources for responding to severe and long-duration nuclear emergencies. Elements of an effective communications plan with the affected populations are discussed separately in Section 6.5.3.
6.5.2 Principles for Formulating Protective Actions
The Fukushima Daiichi accident demonstrated that evacuation of populations at risk is problematic if not executed carefully, and it revealed challenges for evacuation of children, the ill, and elderly (see discussion in Section 188.8.131.52). It also revealed that evacuation, when used as a default protective action, is problematic when long-term consequences related to relocation, mental health impacts to the evacuated population, as well as the material impacts such as loss of business or employment are not considered.48
Additionally, the continuing concerns in Japan that dose levels applied for the protection of the population as a whole do not provide sufficient
47 In the United States, the USNRC amended its regulations in 2001 to require that state and local emergency planners consider the use of KI to supplement other protective actions in the case of a general emergency at a nuclear plant (USNRC, 2002c). Since 2002, if a state requests it, the USNRC will provide enough KI for one or two daily doses to the population within the 10-mile EPZ (USNRC, 2002b). The USNRC expected to issue further guidance but, in 2002, decided not to pursue that effort (USNRC, 2000a, 2001, 2002a). The USNRC has continued to replenish state stockpiles in accordance with expiration dates. Since the inception of the program in 2001, the USNRC has shipped over 47,000,000 KI tablets to participating states. In 2004, the National Research Council issued a report (NRC, 2004a) in response to a congressional request for “a study to determine what is the most effective and safe way to distribute and administer potassium iodide tablets on a mass scale.” (The congressional request is described in the Public Health Security and Bioterrorism Preparedness and Response Act of 2002. Section 127 is available at http://www.gpo.gov/fdsys/pkg/PLAW-107publ188/pdf/PLAW107publ188.pdf, accessed March 20, 2014.) The National Research Council report noted the need for a strategy whereby local planning agencies could develop geographic boundaries for a KI distribution plan based on site-specific considerations. These geographic boundaries would be decoupled from the planning boundaries of the 10-mile EPZ (NRC, 2004a). The USNRC is currently planning to consider prestaging KI outside the 10-mile EPZ (available at http://www.nrc.gov/reactors/operating/ops-experience/japan-dashboard/priorities.html#tier-03, accessed March 20, 2014).
48 ICRP notes that decision makers need to justify disruptive protective actions from the perspective of the radiation exposure saved but also consider other issues that are beyond the scope of radiation protection (González et al., 2013).
protection to children could also arise during the response to a severe nuclear accident in the United States: PAGs in the United States are generally based on average risks for the total population and do not provide separate guidelines for children.49
In view of the Fukushima Daiichi accident, the nuclear industry and organizations with emergency management responsibilities in the United States should assess the balance of protective actions for offsite populations affected by severe nuclear accidents. The analysis should specifically address protective actions for special populations (children, ill, and elderly) and their caregivers and the long-term impacts of sheltering in place, evacuation, and/or relocation, including social, psychological, and economic impacts. It does not appear that USEPA explicitly informed its analysis in the recent PAG draft manual (USEPA, 2013) with lessons learned from the Fukushima Daiichi accident.50
6.5.3 Plans for Communicating with the Public
The importance of a good communication strategy during a crisis is recognized by U.S. government agencies and internationally. The USNRC, FEMA, and the IAEA, for example, emphasize in their guidelines (Persensky et al., 2004; USNRC, 2011a,c; FEMA, 2013a; IAEA, 2013a) the need to deliver understandable, accurate, and timely information while acknowledging uncertainty when communicating with the public. What needs to be evaluated within the United States is how agencies and organizations
49 There are at least two exceptions: The Food and Drug Administration provides guidance on the threshold for taking KI, which includes a separate limit for children. It also recommends that state and local agencies should consider applying the threshold for children to the entire population to simplify decision making in an emergency. Also, FDA’s Derived Intervention Levels (DILs) include consideration of children in the guidance on food and water interdiction (USFDA, 1998). This same approach is not taken with doses to other organs and therefore does not apply to evacuation or shelter-in-place orders.
50 In a November 13, 2013, conference call with the committee, a representative of the USEPA indicated that the agency had taken the experiences from the Fukushima Daiichi accident into account when updating the PAG manual. However, the committee did not find explicit evidence for this in that draft (USEPA, 2013). The draft manual recommends evacuating an area if the expected dose for the first few days of the accident is within the 10- to 50-mSv range. However, informed by the Katrina and other domestic natural disasters, which showed that emergency evacuation plans in nursing homes and hospitals were inadequate in many parts of the United States (Wise, 2006; Blanchard and Dosa, 2009; OIG, 2012; Fink, 2013), the draft manual notes that sheltering in place may be a preferred protective action for special populations (e.g., the elderly and the sick who are not readily mobile) at projected doses of up to 50 mSv over 4 days. When evacuations are deemed difficult because of weather conditions or other hazards, sheltering in place may be justified for those populations for projected doses up to 100 mSv. Additionally, any decisions about sheltering in place for these vulnerable populations would also apply to their caregivers, typically young and healthy individuals.
with emergency management responsibilities coordinate their efforts to effectively deliver informed and concise messages to the public.
Announcements about radioactive material releases from the Fukushima Daiichi plant triggered public health concerns in the United States, especially on the West Coast and Pacific Islands (Tupin et al., 2012). As suggested elsewhere (Salame-Alfie et al., 2012), the Fukushima Daiichi accident could have been used as a test scenario for how communications among the responders and the public would play out if an accident were to occur in the United States. The U.S. National Response Framework (see Sidebar 6.2) was not followed, so there was no declaration of a lead federal agency for the response, and a Joint Information Center with collocated group of representatives from agencies and organizations with the responsibility to handle public information needs was not established.51
During the accident, authorities in U.S. states received a number of inquiries from members of the public regarding potential health effects from radioactive material releases from the Fukushima Daiichi plant; the safety of milk, water, and food; and need to take KI (Salame-Alfie et al., 2012). Thyroid dose projections for U.S. populations were well below levels that would trigger health concerns52 and therefore were not high enough to meet USEPA guidelines for taking KI. Despite this fact, the U.S. Surgeon General’s office issued a statement indicating that that it was appropriate for West Coast residents to take KI. This statement was viewed as incorrect by many (Fitzgerald et al., 2012; Salame-Alfie et al., 2012).
The USNRC informed the public that “no radiation at harmful levels would reach the United States” (USNRC, 2011e) and the USEPA announced that any radioactivity detected in the United States was “well below any level of public health concern.”53 However, little authoritative information was available about the human health impacts of radiation exposures; as a result, the fear of radiation exposure and public perceptions of exposure risks were not consistent with the messaging from government agencies (Haggerty, 2011; Payne, 2011).
The experience of the United States during the Fukushima Daiichi accident highlights the need to review existing plans for communicating with the public during a nuclear emergency. It is important that such plans deliver clear, timely messages about the status of the emergency and notifications of planned actions; recommendations regarding actions that could be taken by affected individuals; frank discussions of uncertainties and unavailable but necessary information; and clarification or correction
51 However, as stated in the Joint Information Center (JIC) manual (NRT, 2000), the structure of the JIC could be useful in coordinating multiagency events internationally.
52 Atmospheric dispersion of the radioactive materials released from the Fukushima Daiichi plant greatly reduced their concentrations by the time they reached the United States.
of alarming information and rumors originating from various sources.54 These communication capabilities need to span all phases and activities related to an accident.
Communicating with the public about the meaning of radiation dose limits during a nuclear emergency is also important. The public confusion about dose limits that occurred in Japan (see Section 6.4.4) would most likely also occur in the United States: the United States has established a variety of radiation dose and radioactive contamination limits for different purposes; these limits are enforced by different means and different agencies. As seen in Table 6.2, dose standards applicable to the general public in the United States range from 1.0 mSv in 1 year from normal nuclear operations to 100.0 mSv in an emergency. This is a factor of 100 difference. There are additional standards, which are not included in the table, that are specific to individual organs (e.g., the thyroid).
Not all of the public confusion originates from the existence of too many standards; another source of confusion is the lack of a separate standard for children (González et al., 2013). As noted in Section 6.4.4, the concerns in Japan that dose levels applied for the protection of the population as a whole do not provide sufficient protection to children suggests that similar concerns could also arise in the United States.
6.5.4 Decision Making for Recovery
The ongoing offsite response to the Fukushima Daiichi Accident demonstrates that cleanup and resettlement of evacuated populations (collectively described here as “recovery”) are complex processes. Many aspects of recovery, including issuing predetermined protective action criteria, cannot be planned in detail before an accident occurs; indeed, such criteria depend on the accident scenario, its consequences, and stakeholder preferences. However, the current situation in Japan, where about half of the evacuees (WNA, 2014) continue to live in shelters or temporary locations with uncertainty about their future plans, emphasizes the need for the United States to conduct advance planning for recovery from a nuclear plant accident.
54 Reviewing the communication efforts during the 1979 Three Mile Island (TMI) accident could offer useful insights. In that instance, considerable trust was established by government leaders when Harold Denton, Director of the USNRC’s Office of Nuclear Reactor Regulation and President Carter’s personal adviser for the TMI accident, took over as spokesperson. He was well equipped to answer many of the questions that the public has been found to worry about in a crisis: What happened? What is being done about it? What should we do? What is likely to happen next? What is your credible worst-case scenario? What are you doing to prevent it? The USNRC’s Special Inquiry Group tasked to investigate the TMI accident describes Denton as a person who if he does not have the answers, “will be willing to look for them and to share them once they are found” (Rogovin et al., 1980).
TABLE 6.2 Selected U.S. Radiation Dose Guidelines for Members of the Public
|Circumstance or Pathway||Standard (mSv)||Agency|
|Drinking water (per year)||0.04||USEPAa|
|Air effluents (per year)||0.1||USEPAb|
|Decommissioned site (per year)||0.25||USNRCc|
|Normal nuclear operations (per year)||1.0||USNRC/USDOEd|
|Relocation (standard per year after year 1)||5.0||USEPAf|
|Lower evacuation threshold (early-phase nuclear power plant accident—first 4 days)||10.0||USEPAb|
|Relocation (first-year dose)||20.0||USEPAb|
|Upper evacuation threshold (early-phase nuclear power plant accident—first 4 days)||50.0||USEPAb|
|Evacuation with serious adverse external conditions for special populations (during one incident)||100.0||USEPAb|
a40 CFR § 141.66(d); from beta and gamma dose.
b40 CFR §§ 61.92 and 61.102 and 10 CFR § 20.1101(d).
c 10 CFR § 20.1402.
d 10 CFR §§ 20.1301 and 835.208.
e USFDA (2004).
f USEPA (2013).
The 1992 USEPA PAG manual (USEPA, 1992) did not address recovery following a nuclear plant accident. USEPA’s recently updated PAG manual (USEPA, 2013), which is still labeled as a draft, minimally addresses recovery. It recommends that resettlement criteria should be established after a contamination event has occurred and notes that the process for establishing such criteria could take months to years. The draft PAG manual also recommends that the process to determine acceptable criteria for a given community should include input from community members and other stakeholders. However, no guidance is given on how to address stakeholder concerns that would likely arise in a Fukushima Daiichi-scale accident and how they might be minimized.
The USEPA draft PAG also does not provide specific recommendations for dose thresholds for long-term cleanup. It references the 1 in 10,000 to 1 in 1,000,000 acceptable lifetime risk criterion for cancer incidence, a range that is generally used for cleanup of contaminated sites under the Comprehensive Environmental Response, Compensation, and Liability Act (CER-
CLA) and the USNRC’s process for decommissioning and decontamination of nuclear facilities. Assuming that the risk of developing cancer increases in proportion to dose received with no threshold (i.e., the LNT model), this risk range translates to an approximate dose to the whole body of 0.009-0.9 mSv over a lifetime.55 The Fukushima Daiichi accident recovery has demonstrated that attaining cleanup goals in this range (i.e., a small fraction of the radiation dose received from natural background in a lifetime) may be impractical when contaminated areas are large.
International radiation protection agencies, such as the ICRP and IAEA, advocate for the principle of optimization when it comes to protection of populations living in an existing exposure situation such as in the areas contaminated by radioactive material releases from the Fukushima Daiichi plant (ICRP, 2007). This approach is a departure from conventional cleanup guidelines under CERCLA or decontamination of nuclear sites, both of which are based on either radiation dose or health-risk levels. The intent of these international recommendations is to take into account not only risk of developing cancer in the future, but also competing factors, for example, the local economy, future land use, cleanup options, and, ultimately, public acceptance. The NCRP, consistent with the ICRP recommendations, is currently (June 2014) finalizing a study that establishes the framework of an approach to optimizing decision making for recovery.56
Deciding on recovery strategies for severe nuclear accidents and their implementation should be part of the U.S. government’s advance planning. The U.S. government should be able to develop and articulate guidance for state and local authorities in dealing with radiation contamination recovery. Issues for which needed policies and decision criteria are required include resettlement and decontamination, including disposal, reduction of volume, or storage of removed contaminated materials. In cases where resettlement may not be desirable, policies will also need to be developed for redirection of (and assistance to) evacuated populations to alternative permanent homes in new locations.
55 The committee derived this accumulated dose range estimate as follows: Using the LNT model, the risk of cancer incidence (all cancers) for a dose equal to 1 mSv/yr over a lifetime is 621 per 100,000 for men and 1,019 per 100,000 for women (NRC, 2006a, Table 12D-3). Assuming a 50:50 gender ratio within a population, the risk for the population as a whole is 820 per 100,000 or else 8,200 per 1,000,000. For USEPA’s reference to the 1 in 10,000 to 1 in 1,000,000 acceptable lifetime risk criteria for cancer incidence, the effective dose would be 0.012 mSv/yr to 0.00012 mSv/yr. Assuming a 75-year average life span, the lifetime dose would be equal to 0.009 to 0.9 mSv over a lifetime. For comparison, the annual average effective dose from background radiation to populations in the United States is 3.1 mSv annually (NCRP, 2009).
56 Presentation by S. Y. Chen, http://www.ncrponline.org/Annual_Mtgs/2014_Ann_Mtg/PROGRAM_2-10.pdf. Last accessed March 20, 2014. The study will be published as NCRP Report No. 175.