Skip to Navigation
Skip to Content
Share this page


Welcome to PSR's Environmental Health Policy Institute, where we ask questions -- then we ask the experts to answer them. Join us as physicians, health professionals, and environmental health experts share their ideas, inspiration, and analysis about toxic chemicals and environmental health policy.


More Topics »

Evacuation in Case of Nuclear Reactor Accident: Feasible?

Posted on October 31, 2012

By Andrew S. Kanter, MD MPH

Accidents can and do happen at nuclear reactors.  The March 2011 nuclear accident in Fukushima, Japan provides important lessons regarding the danger to public safety and the need for evacuation zones around nuclear reactors that are appropriate given the populations who are at risk of injury and death. 

Immediately after the Fukushima accident, the Japanese government evacuated a three-kilometer zone around the plant.  That zone was then expanded to 10 kilometers and later to 20 kilometers, with a stay-indoors zone from 20 to 30 kilometers.  Radiation detection information showed that significant radiation was detected well outside the official 20-kilometer zone, with aerial monitoring by unmanned drone aircraft detecting a swath of elevated radiation northwest from the plant out to about 25 miles [1].  Most areas around Fukushima were apparently spared a more significant exposure because much of the time the prevailing wind was blowing out to sea (it is thought that up to 80% of the radiation went out over the sea instead of over land).  Despite this, radiation measurements of soil samples taken as far away as 50 kilometers from the site showed levels of cesium-137 so high, they exceeded the cut-off used for determining the long-term evacuation zone used around the devastated Chernobyl plant in the then-Soviet Union.

Clearly, the original evacuation zones around the Fukushima reactors, like the 10-mile evacuation zones currently mandated in the United States [2], are insufficient.  In fact the Chairman of the Nuclear Regulatory Commission, Gregory Kaczko, shortly after the Fukushima accident announced a 50-mile evacuation for all Americans near the plant.  However, his recommendation was not accepted by the Japanese government for its own people, nor has it been adopted as standard evacuation planning for U.S. nuclear reactors.

Using simulation software provided by the U.S. government [3] Physicians for Social Responsibility (PSR) performed an analysis of potential radiation levels from a hypothetical nuclear reactor accident near a major metropolitan area – in this case, the Braidwood reactor outside of Chicago. The simulation modeled a loss-of-coolant accident with exposure of the reactor core, a containment breach, and release of the reactor’s superheated radioactive fuel into the air. This is a worse-case scenario, but highlights the dangers from a single reactor breach compared to the multiple reactor meltdowns which occurred in Fukushima.

The simulation showed that the resulting plume of radioactive materials would extend north from the reactor itself to the northern edges of metropolitan Chicago, and east into Indiana and Michigan.  The area that would require evacuation or other protective measures would include the majority of the City of Chicago (population roughly 2.7 million), extending east to Gary and South Bend, Indiana. However, experience from Hurricane Katrina indicates that even one million people with several days’ notice could not be adequately evacuated.

In the case of such an accident, the population would be exposed to different levels of radiation depending on distance from the reactor, duration of exposure (for this simulation, it was assumed that the exposure would continue for one week), and the wind pattern. The simulation indicated that more than 7.5 million people would be exposed to radiation, receiving greater than the maximum allowed annual population dose.  Of that 7.5 million, 4.6 million would receive a dose equivalent of the maximum allowable occupational exposure for one year. More than 200,000 would receive high enough doses to develop radiation sickness, and 20,000 might receive a lethal dose (LD 50), according to our projections. 

The number of acutely ill people in this scenario would overwhelm all available care facilities; about 113 hospitals, including two Veterans Administration hospitals, would fall within the occupational exposure zone, affecting more than 32,000 hospital beds. Nearly 20,000 physicians in five counties exposed to the plume would receive greater than occupational maximums for radiation exposure.

First responders, including firefighters, would also be injured. The closest firefighters would either suffer lethal doses or suffer from radiation sickness and be unlikely to provide a sustained response to the emergency. Another 10,500 firefighters in 355 other departments would exceed occupational exposures from the plume itself and would be unavailable to respond within the highly contaminated area. Police departments also would be hard hit in the closest towns, with an estimated 38 police officers receiving potentially lethal doses of radiation.

Even the current sheltering-in-place recommendation for vulnerable populations who might be harmed by the evacuation does not actually remove the need for eventual evacuation. Experience from Chernobyl and Fukushima shows that exposure to radiation from the ground, water and food eventually forces evacuation of even those who initially were told to shelter in place.

In short, the authorities and health care system would not be able to properly protect the health of all the people and particularly vulnerable populations who would need to be moved in the case of such an accident, let alone the massive number of injured, potentially injured and symptomatic victims. It is therefore likely that many environmental justice populations would receive a disproportionate health burden in the case of a reactor accident. [4]

Dr. Edwin Lyman of the Union of Concerned Scientists (UCS) performed a similar study, proposing a core meltdown at the Indian Point nuclear power plant, located 35 miles north of New York City. That scenario showed an even higher death toll and greater destruction.[5] In that study, a meltdown at the Indian Point power plant could result in 44,000 people dying from radiation poisoning within a year and 518,000 cancer deaths over time. Millions of people in the greater New York City area would have to be permanently relocated, because the resulting contamination would leave huge geographic areas uninhabitable for many years or decades. Economic losses, according to the UCS study, could be from $500 billion to $2 trillion.

Taking this information into consideration, how feasible would evacuation be in the case of such an accident?  PSR estimates that, if we were to utilize the appropriate 50-mile evacuation zone around the Indian Point nuclear reactor, over 17 million people would fall within the evacuation zone. The Calvert Cliffs nuclear plant located outside Washington, DC would require 3.1 million people to be evacuated.  In both cases, there is no conceivable way those people could be evacuated safely in the case of a serious accident at the reactor.  Yet those cases are not unusual.  Taking into account all the nuclear reactors in the U.S. and along the U.S.-Canada border, over 111 million people – one-third of the population of the U.S. – live within 50 miles of a nuclear reactor.  We need to understand that evacuation in the case of a nuclear reactor meltdown is not feasible in many cases, and does not provide the prospect of safety for millions of Americans.  Any discussion about the proper planning for a nuclear reactor accident must take into account the realistic possibility of evacuation and the likely consequences for the health of the of people (including vulnerable populations) living within 50 miles.

[1] The Situation in Japan, the US Department of Energy

[2] Emergency Plans, the US Nuclear Regulatory Commission

[3] Hazard Prediction and Assessment Capability (HPAC) from the Defense Threat Reduction Agency and the Consequence Assessment Tool Set (CATS) from the Federal Emergency Management Association).

[4] Clearwater contention to the NRC Atomic Licensing Board

[5] Lyman ES. Chernobyl on the Hudson? Union of Concerned Scientists, September 2004.


Comments closed.