The Windscale fire of 10 October 1957 was the worst nuclear accident in British history and is also an accident rated Level 5 on the International Nuclear Event Scale.1 The Windscale reactors, where the accident occurred, were built quickly by the British nuclear weapons project.2 Windscale Pile 1 went into operation in October 1950, and Pile 2 in June 1951.3 The Windscale fire was an accident in which a fire broke out in the core of Pile 1, built at Windscale in Cumberland (now Sellafield, Cumbria), and residual radioactive material was released to the outside. At this time, iodine-131, which causes thyroid cancer, was released, and after the accident an estimated 240 people developed thyroid cancer.4 No one was evacuated because of this accident, but milk within 500 square kilometres was disposed of within one month.
Windscale reactors
After World War II ended, the British government judged that it could no longer wait to join the new arms race, and it rapidly advanced its nuclear weapons project.
As part of the nuclear weapons project, Windscale Piles 1 and 2 were built in the small village of Seascale in Cumberland. The two reactors were located only tens of metres apart, and each was housed inside a huge concrete building. These reactors used graphite as a moderator and air as a coolant to remove heat generated by the fuel. Cold air was drawn in by several huge fans, and the hot air that had cooled the reactor exited on the opposite side of the core and was discharged through a chimney. At the top of the chimney was a filter installed on the orders of John Cockcroft. Engineers thought this filter was nearly useless and wasted money and time, and they struggled to lift it to the top of the 120-metre chimney at the end of its construction. For this reason they gave the filter the name "Cockcroft's folly," but this Cockcroft's folly prevented the Windscale fire disaster from becoming a catastrophe.
Core design
The reactor was built with a solid graphite core, and uranium and isotope cartridges were inserted through horizontal channels bored in the graphite, producing plutonium along with isotopes produced by neutron irradiation. Fuel and isotope cartridges were inserted at the "charge face" at the front of the core, and spent fuel was pushed out at the "discharge face." Water-flowing channels installed on the discharge face cooled the cartridges and moved them toward the reprocessing stage.
Fuel cartridges
Inside aluminium cartridges was unenriched metallic uranium, and the cartridges had cooling fins to cool the fuel during reprocessing. To make the plutonium usable as weapons-grade material, heavy plutonium isotopes (such as 240Pu and 241Pu) were burned up in the reactor to reduce their proportion.
Isotope cartridges
Metal cartridges were used to produce isotopes by irradiating them with neutrons. Below is a list of the isotopes to be produced and the materials used to produce them. Among the isotopes produced here, polonium-210 was released in the accident and contributed the most to the collective radiation exposure of the general public.5
Wigner energy
When building the Windscale reactors, Britain had some knowledge of the reactions that occur when graphite absorbs neutrons, unlike the United States or the Soviet Union. The Hungarian-American physicist Eugene Wigner discovered that when graphite is irradiated with neutrons, the crystals on its surface are disrupted, storing neutron energy as potential energy. This stored energy can spontaneously be released as sudden intense heat. When Windscale Pile 2 had been completed and was operating, the core temperature rose abnormally, and this was presumed to be due to the release of Wigner energy. British scientists who were concerned about this began looking for a way to safely release the Wigner energy. The only solution was an annealing process: they discovered that raising the temperature of the graphite core to 250 degrees Celsius would return the disrupted atoms to a crystalline state.6 The annealing process prevented the accumulation of Wigner energy, but apart from the monitoring devices, auxiliary systems such as the reactor and cooling equipment had not been designed for this at all. Over time, the annealing process was changed slightly and became more difficult, and some pockets of Wigner energy remained even after the annealing process.
It was long believed that the graphite moderator burned the most during the accident, but in fact the uranium fuel burned more. A 2005 investigation revealed that the parts of the graphite that were partially damaged were around the burned fuel cartridges.7 Because the annealing process was not originally planned, the thermocouples installed in the reactor monitored only normal operation and did not monitor the annealing process. As a result, the areas of high temperature in the core were not known. Also, unlike the uranium oxide fuel used today, the metallic uranium used at that time ignited easily when it met oxygen. Furthermore, when the fire broke out, the air introduced to cool the reactor passed through the filter while carrying radioactive isotopes and was released into the environment.
Isotope cartridges
The following materials were placed in metal cartridges to produce radioisotopes through neutron irradiation. Among these, the release of polonium-210 helped increase the radiation exposure of the general public.8
Accident
Change of purpose
In 1958, Britain needed to demonstrate technical parity in order to conclude a nuclear weapons agreement with the United States. After the nuclear tests were successfully completed, the United States developed and succeeded in building a hydrogen bomb using tritium. Winston Churchill then declared that Britain could build a hydrogen bomb and handed scientists a tight schedule to achieve this. Britain had no factory to produce tritium, so it decided to use the Windscale reactor. Accordingly, it decided to insert lithium-6 into the reactor and produce tritium from the reactor's neutrons. Because high neutron flux was indispensable for producing plutonium, the cooling fins of the aluminium fuel cartridges, which removed heat, were reduced in size. As a result, the aluminium absorbed fewer neutrons. Because of the tritium production specifications that exceeded what the Windscale reactor was designed for, tritium was produced at reduced safety margins. Afterward, the tritium was successfully extracted from Windscale Pile 1, and while the heat problem was set aside, when the reactor's temperature rose above the design level, scientists changed the thermal distribution of the core, and as a result a hot spot developed in Pile 1. Furthermore, the thermocouples measuring the reactor's heat were not installed according to the changed thermal distribution, so they could not measure the temperature at the reactor's hot spot and instead produced incorrect readings.
Ignition
On 7 October 1957, a small electric current was passed through the cooling fans for the annealing process of Windscale Pile 1, and the reactor was kept at low power. The following day, after the annealing process was completed, the power was raised. During the annealing process, control rods had been inserted into the reactor, but they were withdrawn too early, before the Wigner energy had properly spread through the core. The operators decided to withdraw the control rods again and reheat the reactor to complete the annealing process. It is not well understood, but the second heating is presumed to have caused the fire. The official report concluded that the uranium cartridges had been exposed, oxidized, and overheated, which started the fire; however, a more recent report suggests that the magnesium/lithium cartridges caused the fire. In any case, a smooth temperature rise was observed, during which the Wigner energy was released.
Early in the morning of 10 October, something strange happened. The reactor temperature gradually decreased after releasing the Wigner energy, but the thermocouple readings were ambiguous, and one thermocouple showed the temperature rising. To cool the heat, the airflow was increased. As a result, a large amount of oxygen entered and a fire broke out, and many radioactive materials were carried in the smoke to the chimney. Afterward, the control room found that the radiation monitor readings were reaching a peak. Following the guidelines that had been written earlier, the supervisor declared an emergency.
Fire
The operators tried to determine the reactor's condition using remote measurement devices, but were hindered. Tom Hughes, the deputy reactor manager, suggested directly observing the reactor, so he and another operator put on protective clothing and went to the reactor's charge opening. The measuring plugs of the fuel channels had come out, and nearby thermocouples recorded high temperatures, and they saw the fuel glowing red.
It was now established that the reactor was on fire, and the fire continued for 48 hours. Tom Tuhoy, the reactor manager,9 wearing protective clothing and breathing apparatus, inspected the discharge face from 24.384 metres (80 feet) above the reactor building. He reported in his report that a grim red glow was visible behind the reactor and also reaching the rear containment vessel. In order to prevent the concrete from collapsing due to the high temperatures, he moved back and forth several times during the fire between the discharge face and the rear of the concrete containment vessel.10
Initial fire-fighting attempts
The operators were not certain that a fire had broken out. At first they ran the cooling fans at maximum to extinguish the fire, but this only fanned the flames. Tom Hughes and his colleagues removed the cartridges that were still intact to create a firebreak, and Tom Tuhoy proposed pushing some melted cartridges out into the water channel with a club. However, the fuel rods would not fall into the water channel no matter how much force was applied. The rods used to remove them became hot at the ends, and the rods began to drop molten metal. Hughes realised that this was molten radioactive uranium, and that this caused serious radiation problems for the charge hoist itself.
"It [the exposed fuel channel] was white," said Hughes and the colleague who operated the hoist. "It was really burning white. Nobody, I mean nobody, could believe how hot it really could be."
Carbon dioxide injection
Afterward, the operators injected carbon dioxide to extinguish the fire. At the newly built Calder Hall reactor, 25 tonnes of liquid carbon dioxide had just arrived, and this was improvised into Windscale Pile 1. However, the amount was insufficient to extinguish the fire, because the fire was so intense that it separated oxygen from the carbon dioxide and burned it.
"So we put in carbon dioxide," Hughes said. "And we had a pitifully small carbon dioxide tube, and we found that this offered no hope at all."
Water injection
On the morning of Friday, 11 October, the fire grew worse, and the 11 tonnes of uranium glowed. The temperature rose further (one thermocouple recorded 1300 degrees), and there was even a risk that the reactor's biological shield would collapse. In this crisis, the operators injected water. This was dangerous, because molten metal oxides could come into contact with water, converting water into oxygen and hydrogen, which could either fly off or explode and blow away the weakened structure. With no other options, the operators carried out this plan. More than twelve fire hoses were brought in, nozzles broke, and hose lines were connected to rods a few metres from the heart of the fire, where the improvised fuel channels were. Tuhoy went back up onto the reactor shielding and ordered water to be poured in, and as the pressure increased, he checked for signs of hydrogen explosion. However, water had difficulty extinguishing the fire, so additional measures were needed.
Air cutoff
Tom Tuhoy then had everyone except himself leave the reactor building, and he had the fire chief shut off all ventilation and cooling systems leading to the reactor. Tuhoy went up several times and observed the slowly dying fire. On one of his trips to investigate, he found a metal hook and an inspection plate on the opposite side of the core that had fallen out. He reported that this was caused by the fire trying to draw air from wherever it could.
"At that point I knew the fire was trying to draw air out of the chimney, and that it would keep the fire going," he pointed this out in a later interview.
Finally he removed the inspection plate and welcomed the dying fire.
"After the flames had gone first, the heated material began to die," he explained. "I went back and forth several times until the fire was completely out. I was on the hopeful side," he added, "but if you went up to the front of the extinguished reactor core, you would receive a lot of radiation."
Water was kept running for 24 hours until the reactor had completely cooled.
The reactor tank itself was sealed after the accident and still contains about 15 tonnes of uranium fuel. This was because it was thought that if the remaining fuel became unstable, pyrophoric uranium hydride could ignite again.11 However, in subsequent research needed for decontamination after the accident, this possibility was ruled out.7 Windscale Pile 1 has not been scheduled for final decontamination work until 2037.
After the accident
Release of radioactive material
As a result of this accident, radioactive material was spread over Britain and Europe.4 The Windscale fire released 740 terabecquerels (20,000 curies) of iodine-131, 22 terabecquerels (594 curies) of caesium-137, 12,000 terabecquerels (324,000 curies) of xenon-131, and other radionuclides.12 Data compiled later suggest that more radioactive material may have spread across Britain and the world.4
Health
At that time, iodine-131 was released because of the accident. Iodine-131 has a half-life of only eight days, but it is absorbed and stored in the human thyroid gland. For this reason, iodine-131 can cause thyroid cancer. As a result, 200 cases of thyroid cancer were estimated to have occurred, but this has recently risen to 240 cases.4
Because of this accident, no one in the surrounding area was evacuated, but milk was contaminated to a dangerous degree. For a full month, milk within 500 square kilometres was diluted 1000 times and then dumped into the Irish Sea.
Recovery work
The reactor itself was not recovered. Fuel rods that could be removed were removed, and the biological shield was sealed and kept in place. It is estimated that there are about 6,700 fuel cartridges as well as 1,700 isotope cartridges. The reactor remains warm due to ongoing radioisotope activity caused by the fire.13 Windscale Pile 2 was not damaged, but it was judged too dangerous to use, so its operation was stopped. No air-cooled reactors have been built since then.
Inquiry
An inquiry into this incident was held under Sir William Penney from 17 to 25 October 1957. The report on the incident (the Penney Report) was approved by the United Kingdom Atomic Energy Authority and was approved by Parliament as a government white paper in November 1957. This report was declassified in 1988 in the British National Archives, and a revised copy based on the existing recorded transcript was made public.1415
Penney wrote the report on 26 October, which was 16 days after the fire broke out.16 He reached four conclusions:
- The main cause of the accident was the second heating of the reactor on 8 October, which was too soon.
- Regarding the response to the accident, he found that it was "prompt and efficient, and the people concerned devoted themselves to their work."
- Regarding the consequences of the accident, he wrote that "no immediate harm was seen to the public or to Windscale workers." However, the report was very critical of technical and organisational deficiencies.
- He wrote that the organisation needed to be changed, that clearer responsibility for health and safety was needed, and that a more detailed technical assessment of radiation dose limits was required.
The Windscale site was decontaminated and is still in use. Part of the site was renamed the Sellafield nuclear complex and transferred to British Nuclear Fuels Limited, but Sellafield's main work at present is the decommissioning of reactors.
Comparison with other accidents
The amount of radioactivity from the Windscale fire was avoided as the worst in comparison with the Chernobyl accident, but generally this fire is still regarded as the worst nuclear accident until the Three Mile Island nuclear power plant accident occurred in 1979. Epidemiologically, no cancer patients resulted from the Three Mile Island nuclear power plant accident, and this accident is also estimated to have produced no more than one immediate cancer patient, whereas at Chernobyl casualties occurred immediately.17
Also, while Three Mile Island, Chernobyl, and Fukushima involved commercial reactors, the Windscale reactors were purely military-purpose reactors.
In the case of other military reactor accidents, several people were killed: at SL-1 in Idaho, three operators died, and in 1946 at Los Alamos National Laboratory, Louis Slotin died in a criticality accident.
The Windscale accident occurred at the same time as the Kyshtym accident, but the Kyshtym accident was not known to the public until after the Cold War.
Further reading
- 'The Windscale reactor accident--50 years on' R. Wakeford. Journal of Radiological Protection 2007 Sep vol. 27(3) pp211-5. Epub Aug 29,2007
- "Windscale fallout blew right across Europe", Rob Edwards. New Scientist, October 6, 2007.
- Windscale, 1957: Anatomy of a Nuclear Accident, Lorna Arnold. New York : St. Martin's Press 1992
- "Chernobyl: worst but not first", Walter C. Patterson. Bulletin of the Atomic Scientists, August/September 1986.
- 'Secrets of the Windscale fire revealed', F. Pearce. New Scientist vol 99 29 September 1983 p. 911
- 'Windscale; increased cancer incidence alleged', T. Beardsley. Nature vol 306 Issue 5938 Nov 3 1983 p. 5
- "Accident at Windscale No.1 Pile on 10 October 1957". Cmnd. 302. (H.M.S.O., 1958).
- "Accident at Windscale: World's First Atomic Alarm", Hartley Howe. Popular Science, October 1958, Vol. 173, No. 4.
- "An Assessment of the Radiological Impact of the Windscale Reactor Fire", M.J. Crick, G.S. Linsley. NRPB Reports, Oct. 1957, Nov. 1982.
- An airborne radiometric survey of the Windscale area, October 19–22nd, 1957. A.E.R.E. reports, no. R2890. (Atomic Energy Research Establishment).
- The deposition of strontium 89 and strontium 90 on agricultural land and their entry into milk after the reactor accident at Windscale in October, 1957. A.H.S.B. (United Kingdom Atomic Energy Authority).
- 'Accident at Windscale' British Medical Journal 16 Nov 1957;2 (5054) pp 1166-8.
각주