The Windscale Piles (centre and right) in 1985
The Windscale fire of 10 October 1957 was the worst nuclear accident in Great Britain's history, ranked in severity at level 5 on the 7-point International Nuclear Event Scale. The two piles had been hurriedly built as part of the British atomic bomb project. Windscale Pile No. 1 was operational in October 1950 followed by Pile No. 2 in June 1951. The accident occurred when the core of the Unit 1 nuclear reactor at Windscale, Cumberland (now Sellafield, Cumbria) caught fire, releasing substantial amounts of radioactive contamination into the surrounding area. Of particular concern at the time was the radioactive isotope iodine-131, which may lead to cancer of the thyroid, and it has been estimated that the incident caused 240 additional cancer cases. No one was evacuated from the surrounding area, but there was concern that milk might be dangerously contaminated. Milk from about 500 km2 of nearby countryside was diluted and destroyed for about a month. A 2010 study of workers directly involved in the cleanup found no significant long term health effects from their involvement.
The reactors were built in a short time near the tiny village of Seascale, Cumberland, and were known as Windscale Pile 1 and Windscale Pile 2, housed in large, concrete buildings a few hundred feet apart. The reactors were graphite-moderated and air-cooled. Because nuclear fission produces large amounts of heat, it was necessary to cool the reactor cores by blowing air through channels in the graphite. Cool air was taken in by a battery of large fans, hot air was then exhausted out of the back of the core and up the chimney. Filters were added late into construction at the insistence of Sir John Cockcroft and these were housed in galleries at the very top of the discharge stacks. They were deemed unnecessary, a waste of money and time and presented something of an engineering headache, being added very late in construction in large concrete houses at the top of the 400-ft (120 m) chimneys. Due to this, they were known as "Cockcroft's Folly" by workers and engineers. As it was, "Cockcroft's Folly" probably prevented a disaster from becoming a catastrophe.
The reactors were built of a solid graphite core, with horizontal channels through which cartridges of uranium and isotope cartridges could be passed, to expose the isotope cartridges to neutron radiation from the uranium and produce plutonium and radioisotopes, respectively. Fuel and isotopes were fed into the channels in the front of the reactor, the "charge face", and spent fuel was then pushed all the way through the core and out of the back—the "discharge face"—into a water duct for initial cooling prior to retrieval and processing to extract the plutonium.
Unenriched uranium metal in aluminium cartridges with fins to improve cooling was used for the production of plutonium. As this plutonium was intended for weapons purposes, the burn-up of the fuel would have been kept low to reduce production of the heavier plutonium isotopes (240Pu, 241Pu etc.).
The following substances were placed inside metal cartridges and subjected to neutron irradiation to create radioisotopes. Both the target material and some of the product isotopes are listed below. Of these, the polonium-210 release made the most significant contribution to the collective dose on the general population.
- Lithium-magnesium alloy: tritium
- Aluminium nitride: carbon-14
- Potassium chloride: chlorine-36
- Cobalt: cobalt-60
- Thulium: thulium-170
- Thallium: thallium-204
- Bismuth oxide: polonium-210
- Thorium: uranium-233
During the accident, uranium fuel caught fire — not the graphite moderator as is widely assumed. A 2005 inspection showed that graphite damage was localised around burning fuel elements. The annealing phases were not part of the original plan, so thermocouples were placed in positions in the reactor to monitor normal operations, but not to monitor the annealing process. This allowed unknown hot spots to form. The reactor's metallic uranium fuel could easily burn in the presence of oxygen, unlike the uranium dioxide used in modern reactors. The direct venting of the cooling air to the atmosphere meant that any radioactive material released by the core which passed through the filters would be released into the environment.
In order for Britain to engage in a nuclear weapons treaty with the USA, it had to demonstrate that it was a technological equal. The Windscale facility was built to produce plutonium for the first British atom bomb. After the successful explosion of the atom bomb, the USA designed and exploded a thermonuclear bomb requiring tritium. Britain did not have any facility to produce tritium and decided to use the Windscale piles. Tritium can be produced in nuclear reactors by neutron activation of lithium-6. Higher neutron fluxes were needed for this than for producing plutonium and it was decided to reduce the size of the cooling fins (totalling approximately 500,000 individual fins) on the aluminium fuel cartridges, thereby reducing the absorption of neutrons by this aluminium. By pushing the first-generation design of the Windscale facility beyond its intended limits, tritium could be produced at the cost of a reduced safety factor. After a first successful production run of tritium in Pile 1, the heat problem was presumed to be negligible and full-scale production began, but by raising the temperature of the reactor beyond the design specifications, the scientists had altered the normal distribution of heat in the core, causing "hot spots" to develop in Pile 1. These spikes of heat went unnoticed by the scientists because the thermocouples used to measure the core temperatures were positioned based on the original heat distribution design and were not measuring the hottest parts of the reactor, leading to falsely optimistic readings.
On 7 October 1957, operators began an annealing cycle for Windscale Pile 1 by switching the cooling fans to low power and stabilizing the reactor at low power. The next day, to carry out the annealing, the operators increased the power to the reactor. When it appeared that the annealing process was taking place, control rods were lowered back into the core to shut down the reactor, but it soon became apparent that the Wigner energy release was not spreading through the core, but dwindling prematurely. The operators withdrew the control rods again to apply a second nuclear heating and complete the annealing process. Because some thermocouples were not in the hottest parts of the core, the operators were not aware that some areas were considerably hotter than others. This, and the second heating, are suspected of having been the deciding factors behind the fire, although the precise cause remains unknown. The official report suggests that a cartridge of uranium ruptured and oxidised causing further overheating and the fire, but a more recent report suggests that it may have been a magnesium/lithium isotope cartridge. All that was visible on the instruments was a gentle increase in temperature, which was to be expected during the Wigner release.
Early in the morning on 10 October, it was suspected that something unusual was going on. The temperature in the core was supposed to gradually fall as Wigner release ended, but the monitoring equipment showed something more ambiguous and one thermocouple indicated that core temperature was instead rising. In an effort to help cool the pile, the airflow was increased. This fed more oxygen to the fire and lifted radioactive materials up the chimney and into the filter galleries. It was then that workers in the control room realised that the radiation monitoring devices which measured activity at the top of the discharge stack were at full scale reading. In accordance with written guidelines, the foreman declared a site emergency.
Operators tried to examine the pile with a remote scanner but it had jammed. Tom Hughes, second in command to the Reactor Manager, suggested examining the reactor personally and so he and another operator went to the charge face of the reactor, clad in protective gear. A fuel channel inspection plug was taken out close to a thermocouple registering high temperatures and it was then that the operators saw that the fuel was red hot.
"An inspection plug was taken out," said Tom Hughes in a later interview, "and we saw, to our complete horror, four channels of fuel glowing bright cherry red."
There was no doubt that the reactor was now on fire, and had been for almost 48 hours. Reactor Manager Tom Tuohy donned full protective equipment and breathing apparatus and scaled the 80 feet to the top of the reactor building, where he stood atop the reactor lid to examine the rear of the reactor, the discharge face. Here he reported a dull red luminescence visible, lighting up the void between the back of the reactor and the rear containment. Red hot fuel cartridges were glowing in the fuel channels on the discharge face. He returned to the reactor upper containment several times throughout the incident, at the height of which a fierce conflagration was raging from the discharge face and playing on the back of the reinforced concrete containment — concrete whose specifications required that it be kept below a certain temperature to prevent its disintegration and collapse.
Operators were unsure what to do about the fire. First, they tried to blow the flames out by putting the fans onto full power and increasing the cooling, but this fanned the flames. Tom Hughes and his colleague had already created a fire break by ejecting some undamaged fuel cartridges from around the blaze and Tom Tuohy suggested trying to eject some from the heart of the fire, by bludgeoning the melted cartridges through the reactor and into the cooling pond behind it with scaffolding poles. This proved impossible and the fuel rods refused to budge, no matter how much force was applied. The poles were withdrawn with their ends red hot and, once, a pole was returned red hot and dripping with molten metal. Hughes knew this had to be molten irradiated uranium and this caused serious radiation problems on the charge hoist itself.
"It [the exposed fuel channel] was white hot," said Hughes' colleague on the charge hoist with him, "it was just white hot. Nobody, I mean, nobody, can believe how hot it could possibly be."
Next, the operators tried to extinguish the fire using carbon dioxide. The new gas-cooled Calder Hall reactors on the site had just received a delivery of 25 tonnes of liquid carbon dioxide and this was rigged up to the charge face of Windscale Pile 1, but there were problems getting it to the fire in useful quantities. The fire was so hot that it stripped the oxygen from what carbon dioxide could be applied.
"So we got this rigged up," Hughes recounted, "and we had this poor little tube of carbon dioxide and I had absolutely no hope it was going to work."
On the morning of Friday 11 October, when the fire was at its worst, eleven tons of uranium were ablaze. Temperatures were becoming extreme (one thermocouple registered 1,300°C) and the biological shield around the stricken reactor was now in severe danger of collapse. Faced with this crisis, the operators decided to use water. This was risky, as molten metal oxidises in contact with water, stripping oxygen from the water molecules and leaving free hydrogen, which could mix with incoming air and explode, tearing open the weakened containment. Faced with a lack of other options, the operators decided to go ahead with the plan. About a dozen fire hoses were hauled to the charge face of the reactor; their nozzles were cut off and the lines themselves connected to scaffolding poles and fed into fuel channels about a metre above the heart of the fire. Tuohy once again hauled himself atop the reactor shielding and ordered the water to be turned on, listening carefully at the inspection holes for any sign of a hydrogen reaction as the pressure was increased. However the water was unsuccessful in extinguishing the fire, requiring further measures to be taken.
Tom Tuohy then ordered everyone out of the reactor building except himself and the Fire Chief in order to shut off all cooling and ventilating air entering the reactor. Tuohy then climbed up several times and reported watching the flames leaping from the discharge face slowly dying away. During one of the inspections, he found that the inspection plates—which were removed with a metal hook to facilitate viewing of the discharge face of the core—were stuck fast. This, he reported, was due to the fire trying to suck air in from wherever it could.
"I have no doubt it was even sucking air in through the chimney at this point to try and maintain itself," he remarked in an interview.
Finally he managed to pull the inspection plate away and was greeted with the sight of the fire dying away.
"First the flames went, then the flames reduced and the glow began to die down," he described, "I went up to check several times until I was satisfied that the fire was out. I did stand to one side, sort of hopefully," he went on to say, "but if you're staring straight at the core of a shut down reactor you're going to get quite a bit of radiation."
Water was kept flowing through the pile for a further 24 hours until it was completely cold.
The reactor tank itself has remained sealed since the accident and still contains about 15 tons of uranium fuel. It was thought that the remaining fuel could still reignite if disturbed, due to the presence of pyrophoric uranium hydride formed in the original water dousing. Subsequent research, conducted as part of the decommissioning process, has ruled out this possibility. The pile is not scheduled for final decommissioning until 2037.
There was a release of radioactive material that spread across the UK and Europe. The fire released an estimated 740 terabecquerels (20,000 curies) of iodine-131, as well as 22 TBq (594 curies) of caesium-137 and 12,000 TBq (324,000 curies) of xenon-133, among other radionuclides. Later reworking of contamination data has shown national and international contamination may have been higher than previously estimated. For comparison, the 1986 Chernobyl explosion released approximately 1,760,000 TBq of iodine-131; 79,500 TBq caesium-137; 6,500,000 TBq xenon-133; 80,000 TBq strontium-90; and 6100 TBq plutonium, along with about a dozen other radionuclides in large amounts. The Three Mile Island accident in 1979 released 25 times more xenon-135 than Windscale, but much less iodine, caesium and strontium. Preliminary estimates of atmospheric releases by the Fukushima Daiichi nuclear disaster indicate a broadly similar levels to Chernobyl, well above the Windscale fire releases.
|Three Mile Island|
Of particular concern at the time was the radioactive isotope iodine-131, which has a half-life of only 8 days but is taken up by the human body and stored in the thyroid. As a result, consumption of iodine-131 often leads to cancer of the thyroid. It had previously been estimated that the incident could have caused 200 additional cancer cases, although in 2007 the figure was revised upwards to an estimated 240.
No one was evacuated from the surrounding area, but there was concern that milk might be dangerously contaminated. Milk from about 500 km2 of nearby countryside was destroyed (diluted a thousandfold and dumped in the Irish Sea) for about a month. A 2010 study of workers directly involved in the cleanup—and thus expected to have seen the highest exposure rates—found no significant long term health effects from their involvement.
The reactor was unsalvageable; where possible, the fuel rods were removed, and the reactor bioshield was sealed and left intact. Approximately 6,700 fire-damaged fuel elements and 1,700 fire-damaged isotope cartridges remain in the pile. The damaged reactor core was still slightly warm as a result of continuing nuclear reactions. Windscale Pile 2, though undamaged by the fire, was considered too unsafe for continued use. It was shut down shortly afterward. No air-cooled reactors have been built since. The final removal of fuel from the damaged reactor was scheduled to begin in 2008 and continue for a further four years.
The Board of Inquiry met under the chairmanship of Sir William Penney from 17–25 October 1957. Its report (the "Penney Report") was submitted to the Chairman of the United Kingdom Atomic Energy Authority and formed the basis of the Government White Paper submitted to Parliament in November 1957. The report itself was released at the Public Record Office in January 1988. In 1989 a revised transcript was released, following work to improve the transcription of the original recordings.
Penney reported on 26 October 1957, 16 days after the fire was extinguished and reached four conclusions:
- The primary cause of the accident had been the second nuclear heating on 8 October, applied too soon and too rapidly.
- Steps taken to deal with the accident, once discovered, were "prompt and efficient and displayed considerable devotion to duty on the part of all concerned".
- Measures taken to deal with the consequences of the accident were adequate and there had been "no immediate damage to health of any of the public or of the workers at Windscale". It was most unlikely that any harmful effects would develop. But the report was very critical of technical and organisational deficiencies.
- A more detailed technical assessment was needed, leading to organisational changes, clearer responsibilities for health and safety, and better definition of radiation dose limits.
The Windscale site was decontaminated and is still in use. Part of the site was later renamed Sellafield after being transferred to BNFL; the whole site is now owned by the Nuclear Decommissioning Authority.
The release of radiation by the Windscale fire was greatly exceeded by the Chernobyl disaster in 1986, but the fire has been described as the worst reactor accident until Three Mile Island in 1979. Epidemiological estimates put the number of additional cancers caused by the Three Mile Island accident at not more than one; only Chernobyl produced immediate casualties.
Three Mile Island was a civilian reactor, and Chernobyl primarily so, both being used for electrical power production. In contrast Windscale was for purely military purposes.
The reactors at Three Mile Island, unlike those at Windscale and Chernobyl, were in buildings designed to contain radioactive materials released by a reactor accident.
Other military reactors have produced immediate, known casualties such as the 1961 incident at the SL-1 plant in Idaho which killed three operators, or the criticality accident which killed Louis Slotin at the Los Alamos National Laboratory in 1946.
The accident at Windscale was also contemporary to the Kyshtym disaster, a more serious accident which happened on 29 September 1957 at the Mayak plant in the Soviet Union, when the failure of the cooling system for a tank storing tens of thousands of tons of dissolved nuclear waste resulted in a non-nuclear explosion.
In 1999, the BBC produced an educational documentary about the fire as a 30-minute episode of "Disaster" (Series 3) entitled The Windscale Fire. It subsequently was released on DVD.
In 2007, the BBC produced another documentary about the accident entitled "Windscale: Britain’s Biggest Nuclear Disaster", which investigates the history of the first British nuclear facility and its role in the development of nuclear weapons. The documentary features interviews with key scientists and plant operators, such as Tom Tuohy, who was the deputy general manager of Windscale. The documentary suggests that the Windscale fire of 1957 - the first fire in any nuclear facility - was caused by the relaxation of safety measures, as a result of pressure from the British government to quickly produce fissile materials for nuclear weapons.