After the 1946 McMahon Act ended British access to American nuclear secrets, Britain decided it had to build atomic bombs alone. Plutonium was cheaper to produce than enriched uranium - so plutonium it would be. The Americans had built their production reactors at Hanford, far from any city, water-cooled with vast volumes of high-purity water. Britain had no equivalent site that met American distance criteria; applying them would have ruled out all of England and Wales and most of Scotland. Christopher Hinton, head of the production effort, considered Loch Morar and even building in Canada (rejected because the reactors would not have belonged to Britain). In the end, the choice came down to a former Royal Ordnance Factory at the coastal hamlet of Low Sellafield. Water was available from Wastwater. A railway siding was already there. Courtaulds had been planning to use it for rayon manufacture but withdrew. The site became Windscale, named after a bluff above the Calder River. A single reactor was costed at £20 million; two could be built for £30-35 million. They built two.

Hinton's engineers concluded an air-cooled reactor would be cheaper, simpler, safer and faster to build than a water-cooled one. R. G. Newell proposed the reactor be encased in a pressure vessel. D. W. Ginns, H. H. Gott and J. L. Dickson designed finned aluminium cartridges around the uranium fuel rods, which would increase heat exchange. Air would be pulled through horizontal channels in the graphite core, up through tall chimneys, by both natural convection and powered blowers. The decision saved 40 per cent on cost and let construction begin without a water purification plant. Construction commenced in September 1947. At its peak the site employed over 5,000 men. Each reactor sat atop a reinforced concrete slab 200 feet wide, 100 feet long and 10 feet thick. The graphite blocks - 2,000 long tons per reactor, in a 25-sided octagonal stack - had to be machined to tolerances of thousandths of an inch in a clean room environment, by workers wearing special clothing. The biological shield around each reactor was concrete seven feet thick, lined with steel. The chimneys rose 410 feet.

During construction, Sir John Cockcroft visited the X-10 Graphite Reactor at Oak Ridge in the United States. He found uranium oxide particles in the surrounding area. With 70,000 fuel cartridges to be loaded into the Windscale Piles, even a small percentage failure rate would mean burst cartridges. Burst cartridges could oxidise into uranium oxide dust and ride the cooling air straight up the chimneys and into the Cumbrian sky. Cockcroft ordered filters. The engineers were unimpressed - the obvious place for filters was at the bottom of a chimney, and the first seventy feet of the Pile No. 1 chimney was already built. They went on top instead. Two hundred long tons of structural steel, brick, concrete and equipment had to be hoisted to the top of 400-foot chimneys. The filters gave the chimneys their distinctive bulge. Workers and engineers mocked them as Cockcroft's Folly. Years later, it turned out the uranium oxide at Oak Ridge had come from a chemical separation plant, not the reactor. Cockcroft's caution had been based on a misdiagnosis. The misdiagnosis saved northern England.

Pile No. 1 went critical in October 1950, Pile No. 2 in June 1951. Tom Tuohy retrieved the first sample of British plutonium on 28 March 1952. Enough Windscale plutonium for an atomic bomb reached the Aldermaston weapons division by August 1952; Britain's first device detonated in the Monte Bello Islands on 3 October. Then the Wigner energy problem emerged. Graphite under sustained neutron bombardment accumulates dislocations in its crystal structure that release as heat - and could release suddenly. The piles had to be deliberately annealed, heated to bleed off the stored energy. The first anneal of Pile No. 2 took place on 9-10 January 1953. Between August 1953 and July 1957, eight anneals were carried out on Pile No. 1 and seven on Pile No. 2. Each one was slightly different, and each became slightly harder than the last. Meanwhile, on 1 March 1955, Winston Churchill committed Britain to a hydrogen bomb. To produce the necessary tritium without building a new reactor, the Pile fuel loads were modified - lithium-magnesium cartridges added, cooling fins reduced, temperatures pushed beyond design. Windscale staff raised concerns. The concerns were brushed aside. Christopher Hinton resigned. The ninth anneal began on 7 October 1957, and three days later, Pile No. 1 was on fire.

Pile No. 1 still contains roughly 6,700 fire-damaged fuel elements and 1,700 damaged isotope cartridges. As of estimates made in 2000, the damaged core held about 4 kg of plutonium-239, 69 g of caesium-137, 37 g each of strontium-90 and yttrium-90, 4.1 g of tritium, and smaller quantities of other radionuclides. The bioshield was sealed. Pile No. 2's chimney was reduced to the height of adjacent buildings in the early 2000s. Pile No. 1's chimney, far more contaminated, has taken much longer; demolition is ongoing. Final decommissioning of the reactor cores is not scheduled until 2037 at the earliest. The Penney Report concluded that the immediate response to the fire had been "prompt and efficient" and showed "considerable devotion to duty." Tom Tuohy, who stood on the reactor lid during the fire and personally directed the response, never received public recognition during his lifetime. He died on 12 March 2008. "They were a shower of bastards," he said of the officials who let the Americans believe the fire had been caused by reckless workers. The Piles were the last air-cooled production reactors anyone built. The lessons of their design and their failure shaped every reactor that came after.