Here is a list of some of the accidents that have occurred in the two spheres of modern nuclear technology: energy production and the military. It is not a complete list, nor is it a list of the most serious nuclear accidents in history. Instead it is an account of various types of nuclear accidents, and a description of some of the most serious accidents within those types. So, for example, the 1975 mishap at Brown's Ferry is included because it illustrates how easy it is for a nuclear accident to occur in an unusual way, while other incidents that resulted in greater actual hazard, such as partial meltdowns at the Chalk River (Canada) and Idaho Falls reactors, are not included because similar incidents are described elsewhere.
AEC = Atomic Energy Commission
DOD = Department of Defense
- Clarfield, Gerard H., and William M. Wiecek. Nuclear America. New York: Harper & Row, 1984.
- DOD. "Narrative Summaries of Accidents Involving US Nuclear Weapons, 1950-80".
- Lapp, Dr. Ralph. Kill and Overkill.
- MIT faculty. The Nuclear Almanac . Ed. Jack Dennis.
- NYT. New York Times.
- Trabalka, John, L. Eyman and Stanley Auerbach. "Analysis of the 1957-58 Soviet Nuclear Accident". Science 209 (1980).
- Medvedev, Zhores. Nuclear Disaster in the Urals.
- Metzger, H. Peter. The Atomic Establishment .
- Newsweek (10/27/75): 51-52.
- Novick, Sheldon. The Careless Atom. 157-66.
- Ognibene, Peter. "Living with the Bomb: How Safe is Nuclear Pollution?" Saturday Review (6/24/78): 8-12
- Patterson. Nuclear Power, 2nd ed.
A test explosion of a hydogen bomb exposes 290 people to radioactive fallout. The explosion contaminates 7,000 sq. miles of islands and ocean, resulting in widespread sickness and one death due to radiation poisoning. (MIT: 286-87)
5/27/57, Kirtland Air Force Base, New Mexico
A B-36 accidentally drops an unarmed nuclear bomb in the New Mexico desert. The bomb's explosive trigger detonates, which obliterates the weapon, scatters debris up to a mile away, and leaves a radioactive crater 25 ft wide and 12 ft deep. (DOD)
9/11/57, Rocky Flats Nuclear Weapons Plant, Colorado
A plutonium fire burns for half a day, consuming up to 45 pounds of plutonium and exposing much of the Denver area to high levels of radioactivity. The official government spokesman says there was "no spread of radioactive contamination of any consequence." Subsequent tests found high levels of enriched uranium at two elementary schools, 12 miles from the site. (NYT 9/9/81: I,12)
Winter 1957-58, Kyshtym, USSR
An accident in the Urals exposes between 25 and 100 square kilometers of populated land to high levels of radioactivity, and an additional 1000 square km. to levels of radiation significantly above normal. Because of Soviet and US censorship no one really knows the cause or extent of the accident, but the most plausible theory entails an explosion of high level waste associated with Soviet military plutonium production. The accident necessitated the permanent evacuation of thousands of people. An oft-disputed eyewitness account of a Soviet scientist and emigre describes the radiation death of hundreds. (Trabalka. Medvedev.)
1/31/58, Overseas Base
A B-47 carrying a nuclear weapon "in strike configuration" crashes on a runway and burns. The contamination is locally contained. (DOD)
1/24/61, Goldsboro, North Carolina
A B-52 on airborn alert begins to breakup, due to stuctural flaws in a wing, and subsequently releases two nuclear bombs. The automatic parachute of one bomb fails, and the bomb breaks apart when it lands. Of the six interlocking safety mechanisms which must be activated in sequence to set off the bomb, five are triggered, leaving only one switch in the way of detonation of the 24 megaton warhead. Additionally, a portion of uranium from the bomb is lost and never recovered. (Lapp. DOD.)
3/14/61, Yuba City, California
A B-52 armed with two nuclear weapons loses cabin pressure in flight, forcing the crew to bail out after steering the plane away from populated areas. The crash tears the weapons from the plane. There is no contamination or explosion. (DOD)
12/5/65, Pacific Ocean
An A-4 aircraft loaded with a nuclear weapon falls off an aircraft carrier. The weapon is lost in the ocean (along with pilot and plane). (DOD)
1/17/66, Palomares, SPAIN
An airborn B-52 carrying four multi-megaton nuclear weapons crashes into a refueling tanker and drops all four weapons. The explosive triggers for two of the weapons detonate, scattering plutonium. The accident requires removal of 1,500 tons of soil and plant life, to be stored in the U.S. A third weapon is lost at sea for three months. (DOD)
1/21/68, Thule, GREENLAND
A B-52 carrying four multi-megaton nuclear weapons crashes on its landing approach. The weapons are destroyed in the fire. 237,000 cubic feet of radioactive ice and debris are removed and stored in the United States. (DOD)
5/11/69, Rocky Flats N. Weapon Plant, Colorado
Fire consumes between one and four thousand pounds of plutonium, spreading radioactive contamination through several small towns, and into metropolitan Denver, 16 miles downwind of the plant. Over 300,000 cubic ft. of radioactive earth is shipped to a waste storage site in Idaho. Had the fire breached the roof of the plant it would have exposed Denver to lethal doses of radiation. (Metzger: 147-50, 288-89. Clarefield: 354-55)
Fall 1969-70, Rocky Flats, N. Weapon Plant, Colorado
Tests conducted in the aftermath of the 1969 fire uncover leaks from 1,405 drums of contaminated waste. Part of the water supply for the local suburbs is contaminated by the waste, which had been illegally stored at the site for ten years. (Ognibene)
8/7/79, Nuclear Fuel Plant, Erwin, TN
A pipe-clog at a nuclear fuel plant releases super-enriched uranium, exposing about 1,000 people to as much as five times the amount of radiation that they would normally receive in a year. The plant manufactures fuel for the US nuclear submarine fleet. In 1979, the plant lost 22 kilograms of highly enriched uranium. (NYT 10/31/80: I,14)
9/19/80, Damascus, Arkansas
A maintenance man working inside a Titan II missile silo drops a wrench, puncturing the fuel tank of the missile. Several hours later the escaping fuel vapors ignite into a fireball that destroys the missile, blows off the 740-ton top of the silo, and sends the nine megaton warhead flying 200 yards through the air. There was no contamination. (DOD)
1988, Rocky Flats Followup (see 1957, 1969, 1969-70)
In 1976 the county health officer, Carl J. Johnson, reports that soil around the plant contains 44 times more plutunium than the government claimed. In 1977 he reports higher-than-average rates of leukemia and cancer among the local people. In 1980 he reports that plant workers have eight times more brain tumors than expected. In 1981 he is fired. (NYT 12/18/88)
Most of the information about U.S. nuclear weapon accidents comes from a 1981 Department of Defense (DOD) news release titled "Narrative Summaries of Accidents Involving U.S. Nuclear Weapons 1950-1980" (updated Jan. 7, 1982). The document, which contains 32 one-page summaries of U.S. nuclear weapon accidents, is available to the public, and the DOD will send a copy of it first class and for free to anyone who requests it. (202-697-4162, ask for "Salatko"). The document only lists declassified accidents.
In addition to the accident accounts, the document provides some helpful technical and historical information. According to the DOD, "Most of the aircraft accidents...occurred during logistic/ferry missions or airborne alert flights by Strategic Air Command (SAC) aircraft. Airborne alert was terminated in 1968..." Accidents, "particularly those at Palomares and Thule", were one of the main reasons for terminating airborne alert (in addition to cost and the developent of the ICBM), which presumably accounts for the great reduction in the number of nuclear weapon accidents after 1968. It was also customary with early models of nuclear weapons to keep a capsule of nuclear materials separate from the weapon as a safety precaution. No nuclear detonation of a weapon with the capsule removed was possible, although even an unarmed weapon contained a quantity of radioactive material, and accidents involving the detonation of such a weapon's high-explosive (TNT) trigger could and did occur (e.g. the 1957 incident at Kirtland Air Force Base). Current weapon design relies on a number of super-redundant safety features to prevent accidental detonation, so keeping nuclear materials separate from the weapon is deemed unnecessary.
The DOD summaries are not always complete. Some cases involving the early weapon designs state that the nuclear capsule was not in place in the weapon and so no nuclear detonation was possible, but there are also many accounts in which no information is given about the weapon's design or safety features, or of the chances of a nuclear detonation. It is annoying to read the account of a fire in a B-47 aircraft which consumed one nuclear warhead, or about the aircraft loaded with a nuclear weapon which fell off an aircraft carrier into the ocean and was never seen again, and not know the status of the weapons involved. Also disconcerting is the DOD's account of the 1961 incident over Goldsboro, NC, as it fails to make any mention of how severely the accident tested the safety mechanisms of the bomb.
Although many early studies predicted that nuclear power would dominate the energy future...the public does not trust it, with most concern focused on reactor safety. The second generation of reactors, which is waiting in the wings for testing funds, offers great hope in this regard....These reactors include high-temperature gas and sodium metal-cooled types that incorporate innovative new designs to maximize reactor safety. In one design, fuel is encased in small balls that will not melt even if an uncontrolled chain reaction occurs.
--Kessler, E. Stephen. Mineral Resources, Economics and the Environment. Macmillan: New York, 1994: 162-163.
10/8/57, Windscale Nuclear Power Plant, ENGLAND
A fire sends a radioactive cloud floating over the countryside, contaminating hundreds of square miles of dairy farmland. The government dumps half a million gallons of contaminated milk into local rivers. In 1983 the British Government estimated that 39 people had died of cancer because of the accident, caused by operator error. (Clarfield: 349. NYT, 4/86)
10/5/66, Fermi Nuclear Reactor Lagoona Beach, Michigan
A zirconium liner comes loose and blocks the coolant flow to portions of the reactor, causing a partial meltdown of fuel elements, and creating a Class I emergency at the reactor. It takes over two years to discover the cause of the accident because of fear that disturbing the partially melted core will set off a radioactive explosion (Detroit is 25 miles away). (Novick)
6/8/73, Hanford Nuclear Power Plant; Richland, WA
Workers discover that a storage tank has been leaking high-level radioactive waste at a rate of 100 gallons per hour for a month and a half. A total of 115,000 gallons of waste are lost. Since Hanford was established in 1944 it has leaked about 500,000 gallons of high-level waste into the earth, most of which does not seem to have spread to groundwater levels or beyond the local area. (MIT: 262-63, 295)
3/22/75, Brown's Ferry Nuclear Power Plant, Alabama
An electrical fire caused by the use of a candle to check for air drafts results in loss of control of two reactors and the breakdown of the emergency core cooling system for one. Ten hours later a makeshift cooling system is improvised, averting a meltdown. (Newsweek)
3/28/79, Three Mile Island, Pennsylvania
A loss of coolant due to malfunctioning equipment and human error causes a meltdown of fuel in reactor #2, releasing a variety of radioactive products into the environment. The contamination of the facility precludes human entrance for sixteen months. (???)
4/26/86, Chernobyl Nuclear Power Plant, Ukraine
Operators commit a series of safety violations, including turning off the emergency core cooling system and overriding the plant's automatic shut-down procedure, which lead to a fire, explosion, and core meltdown. Accounts of the total amount of radiation released into the environment vary, the US claiming an amount equal to that of all the previous nuclear tests and bombs combined, the USSR claiming somewhat less. 135,000 people were displaced by the fallout, all residents within 18 miles of the plant, as well as children from some villages 100 miles distant, were evacuated, and the European Communtity banned all meat and produce from Eastern Europe for several weeks. 31 people die , and 209 suffer serious radiation exposures. The estimate of the number of cancer deaths during the next 70 years due to the accident ranges from 5,000 to 500,000, although the 6,000 to 24,000 range is seen as most probable. (NYT 10/31/81: I,14)
The Atomic Energy Commission (AEC) permits the use of thousands of tons of uranium tailings "in construction materials, for sand traps on golf courses and for children's sandboxes". Over 5,000 private homes and several schools and businesses are built on the radioactive tailings. The AEC claims it has no right to regulate the use of radioactive material in residential construction. In 1970, Colorado Health Officials discovered that, in one elementary school, "the airborne radioactivity exceeded the federal limits permitted in uranium mines." (Metzger: 169-81)
1/3/61 National Reactor Testing Site, Idaho
A small experimental reactor goes out of control and causes a steam explosion. The operators are killed instantly (one being impaled against the reactor ceiling by a control rod). There is no significant contamination of the environment, although the bodies of the men have to be buried in lead-lined coffins sealed in lead-lined burial vaults, and their heads and hands are severed and disposed as high-level radioactive waste. The accident remains unexplained. Some investigators for the Atomic Energy Commission suggested that it was a deliberate suicide-murder, provoked by jealousies within a lover's triangle involving two of the technicians. (Clarfield: 349-50)
7/16/79 Church Rock, New Mexico
A dam of uranium milling wastes collapses, spilling 100 million gallons of radioactive liquid and 1,100 tons of solid waste into a tributary of the Little Colorado River. A few weeks later, hazardous levels of radioactivity and heavy-metal contamination are detected in the groundwater, forty feet below the surface. (MIT: 378)
Linus Pauling: "...estimated that for every 10 megatons of nuclear explosions (atmospheric), 15,000 children will be born with serious physical or mental defects." (Clarfield: 314) The total explosive yield of nuclear test explosions (close to a thousand in number) is roughly 375 megatons, 295 atmospheric. (Clarfield: 249-50)
Ernest Sternglass (physicist): "...by 1969 reached a shocking conclusion: probably 400,000 infants had died...because of medical problems caused by fallout--chiefly lowered resistance to disease and reductions in birth weight." (Clarfield: 359)
Arthur Tamplin (biophysicist, former AEC investigator) found Sternglass' estimate to be 100 times too high, meaning that 4,000 babies had died instead of 400,000. The AEC, not very happy with this refutation, demanded that Tamplin recant, but he refused and was backed up by John Gofman.
The AEC's Biological and Medical Advisory Committee estimated that fallout to date would produce between 2,500 and 13,000 genetic abnormalities." (Clarfield: 214. Patterson: 142)
New England Journal of Medicine: "A significant excess of leukemia deaths occurred in children up to 14 years of age living in Utah between 1959 and 1967. This excess was concentrated in the cohort of children born between 1951-58 (the height of nuclear testing), and was most pronounced in those residing in counties receiving fallout. J.L. Lyon et.al. "Childhood L. Associated w/ Fallout from N. Testing" N.E.J.M., 300 (1979), 397-402.