A look inside the Dounreay Fast Reactor

Dounreay is one of the strangest places in Britain. It’s literally on the edge of things, isolated in bleakly beautiful landscape on the Caithness cliffs at the extreme north of Scotland, where the mainland starts to fray into the islands of the Orkneys, the 1950s high-tech of its buildings and technology make it seem more like something from science fiction than reality: a James Bond villain’s lair or an early Doctor Who view of the future. It seems like a future which didn’t happen, although as we explain in this issue, breeder reactors whose design is partly based on the Dounreay facilities may be about to return to the UK to handle the country’s nuclear legacy.

The decommissioning of the breeder reactors of Dounreay has now reached a fascinating stage: the engineers have inserted nitrogen-cooled robotic video cameras into the first of the breeders to become functional — the Dounreay Fast Reactor, which is housed within the site’s most iconic structure, the ‘golf ball’ dome. The video cameras are providing the first images for over half a century of the internals of the reactor: ghostly pictures of the breeder elements which are jammed in place, their steel swollen by the neutron radiation, a problem unanticipated when the reactor was built.

The Engineer reported on the early days of the Dounreay Breeder project back in 1955, with an article contributed by two of the senior engineers, JM Kendall and TM Fry. In it, they explain the thinking behind the project: that breeding a fissile fuel — plutonium — by exposing non-fissile uranium-238 to the high-energy neutrons emitted by fission of uranium-235 is a more efficient way to use uranium than to separate out the U-235 and discard the rest. ‘The cost of fissile material is measured in thousands of pounds per kilogram,’ they say, ‘so the initial charge represents a considerable capital investment. If it is to be used as a catalyst for deriving electrical power from U238, each kilogram of fissile material must yield hundreds of kilowatts of electricity. We cannot consider the fast reactor an economic source of power until ratings of this order have been demonstrated.’

In the end, economics was the end of Dounreay. The end of the nuclear arms race in the 1980s caused the price of fissile uranium — which had been in demand for making weapons — to fall, and it no longer made sense to breed nuclear fuel.

Building the DFR was always seen as an experiment. ‘No engineering experience is available on which to design a core with the rating, outlet temperature and burn-up required,’ Fry and Kendall said; the best way to proceed was therefore to design the most flexible plant possible. ‘As failures are replaced by better components, it is hoped to demonstrate that a fast reactor can be a sound economic proposition for the production of industrial power.’

Material choice and construction methods were chosen carefully, and based on those used in highly-radioactive chemical plants which used concentrated nitric acid and had to be made leak-proof. The reactor vessel itself — referred to as ‘the pot’ — along with the primary and secondary coolant circuits, were made from stainless steel which was butt-welded and the welds radiographically inspected along almost their entire length. Sodium-potassium alloy (NaK) was chosen for both coolant circuits, as it could handle the heat from the nuclear reaction; the secondary coolant, which transferred the reaction heat across an exchanger into water to raise steam, had to be chemically inert to the primary coolant, so it was decided to use the same material.

The article lists the many safety precautions which were built into the reactor. The golf ball, 135ft in diameter and an inch thick, was designed to contain any fission products which might escape if the reaction vessel were breached. It was also designed to contain the effects of a sodium fire, in the event of a reactor accident which evaporated the coolant. In this event, Kendall and Fry say, all the oxygen in the sphere could be consumed, creating a partial vacuum inside and allowing the atmospheric pressure outside to begin to crush the sphere. ‘The most pessimistic estimate is that, at 5lb per square inch external pressure, a shallow dimple some 18ft in diameter would appear at the top,’ they said. ‘This would not raise stresses in the steel greater than 1dwt per square inch’.

Nowhere in the article do the authors mention decommissioning the reactor at the end of its life — an omission which had made the task at the plant such a difficult one. DFR closed down in 1977; its successor, the Prototype Fast Reactor in a neighbouring building, closed in 1994. You can read more about the project to decommission the site here. Most of the DFR’s fuel, apart from one rod, and almost all of the NaK has now been removed from the reactor and destroyed by reacting it with water vapour in a purpose-built plant inside the sphere, but 977 breeder elements remain. The elements remain radioactive, and the decommissioning engineers are now trying to work out how to remove the swollen, split metal safely.

‘Being able to see inside the reactor for the first time in half a century is a historic moment and a milestone in the DFR decommissioning programme,’ commented DFR senior manager John Smith. ‘The images and information captured are now enabling the Dounreay team to prepare accurate plans for the safest and most efficient approach for the removal of the remaining fuel and disassembly of the vessel’s internal structures.  They are also assisting in the detailed planning now underway for the cleansing and/or destruction of the residual NaK remaining in the pipes and vessels of the DFR.’

Once the reactor has been disassembled and removed, the sphere itself will be demolished; this is scheduled to happen in 2025. At that point, 70 years after Kendall and Fry explained the project, the Caithness landscape will return to its pristine state.