Monumental task

Once the site for the UK’s experimental fast reactor programme, Dounreay is now the location for one of the world’s largest, trickiest, nuclear clean-up jobs. Stuart Nathan reports.

Just about as far north as you can go without falling off mainland Britain, the landscape is bleakly spectacular. Rolling, windswept hills finish abruptly in jagged cliffs, pounded by Atlantic breakers. And sitting on the edge of the cliffs is the famous Golf Ball sphere of Dounreay — the site for the UK’s experimental fast reactor programme, and now the location for one of the world’s largest, trickiest, nuclear decommissioning programmes.

Dounreay has always aroused strong emotions. In the 1950s, it was a symbol of the UK’s ambition to be at the forefront of nuclear technology. After the US government declared data from the Manhattan Project classified, UK physicists began their own experiments on plutonium and teamed up with engineers to build the world’s first electricity-generating fast reactor in just three years.

In the 1960s and 1970s Dounreay’s technological achievements were overshadowed by suspicions about nuclear power, with stories of missing plutonium, unauthorised discharges of radioactive substances, and explosions in waste shafts. In the 1980s and 1990s the site was condemned as a White Elephant. And now, the long task of clean-up is well under way — a job that will take many times longer than it took to build the site.

Fast reactors were partly a response to the Cold War. In the early 1950s the type of uranium that could undergo nuclear fission was in short supply as it was needed for bombs. Fast reactors use a small amount of fissile uranium fuel surrounded by a ‘breeder blanket’ of non-fissile uranium, U-238. The fuel emits ‘fast neutrons’, which are absorbed by the U-238, transforming it into fissile plutonium-239 that can be reprocessed for fuel, replacing the original uranium. Thus, the reactor ‘breeds’ its own fuel. U-238 was cheap, so the hope was that this would produce very cheap electricity.

The US, USSR and UK all worked on fast reactors, but Dounreay was the first to provide energy to a national grid. The Dounreay Fast Reactor (DFR), housed within the Golf Ball, started providing a modest 15MW to the grid in 1961. The reactor was closed down in 1977, and succeeded by the Prototype Fast Reactor (PFR), intended to be the pilot for a commercial version.

Producing 250MW, PFR became a victim of the end of the Cold War. Fast reactors had proved more complicated and expensive to operate than the original estimates and, once the price of fissile uranium began to fall in the 1980s, the government decided that the technology was uneconomic. Fast reactor R&D began to be phased out in 1988, and the PFR closed in 1994.

When the plants were being built, decommissioning was not considered and many consequences of operating the plants were unknown — for example, nobody knew that bombarding steel with neutrons would cause it to swell, leading to some DFR breeder elements becoming stuck in the reactor. Moreover, the site contains several types of hazardous material. The cooling systems used molten alkali metals, sodium and potassium, which react violently with water, producing explosive hydrogen and caustic fumes. Some coolant was in direct contact with the fuel, and is contaminated with radioactive caesium. Plus, the reputation of plutonium is well-deserved.

Then there are the mishaps — radioactive particles on the seabed and beaches, and the notorious shaft filled with largely unclassified waste.

All of the site’s 180 facilities have to go. A new company, Dounreay Site Restoration (DSRL) has been formed as a subsidiary of the United Kingdom Atomic Energy Authority (UKAEA) and, according to programme strategy manager Doug Graham, has set a target of 2025 to reach what he calls an ‘interim end point’.

By then the nuclear facilities will have been decommissioned, dismantled, and the waste packaged for storage. There is a question mark over the fate of the waste — the Scottish government has said it must remain where it was produced and will not participate in a UK deep waste repository project, but DSRL is following the UKAEA policy and assuming that waste will go to a central repository.

Visiting Dounreay is distinctly eerie. While new technology is being developed and deployed for the decommissioning process, most of it is invisible, inside the reactors or concealed behind shielding. The result is a curious time-warp. The DFR control room still has its Bakelite phones, paper chart recorders, heavy levers, and a big red shutdown button.

The chilly, echoing dome of the Golf Ball, entered through heavy steel doors guarding an airlock tunnel, resembles the villain’s lair in a James Bond film, dominated by the reactor in its pit and the massive Goliath crane overhead. And in the PFR building, the mechanical manipulators for handling fuel dangle above viewing windows like the arms of robots from a science fiction dystopia.

But Dounreay also has its new technology. In PFR and DFR, it is being used to handle the liquid metal coolant in the reactors — sodium in PFR, and a mixture of sodium and potassium, known as NaK, in DFR. The metals are being removed from the reactors — where they have been since they were built — and destroyed in purpose-built plants.

Because the metals react so violently with water, engineers have devised a process that disperses them, after the radioactive caesium has been removed, into a concentrated solution of sodium hydroxide. The metals react with the very small amount of water in the solution, which moderates the reaction and reduces the amount of energy it gives off. Further stages in the process convert the hydroxide into brine, which can be discharged into the sea.

Removing the coolant from PFR has taken eight years. The sodium was kept molten at 150°-250°C by circulating NaK through heating coils in the reactor, which allowed most of the 900 tonnes of primary coolant (the sodium in direct contact with the fuel elements) to be pumped out. But the last five tonnes was a challenge, explained alkali metals residue project manager Billy Husband.

‘The internal geometry of the reactor made it difficult to reach,’ he said. ‘We had to design a pumphead with an integral camera and lighting system, which we could manoeuvre remotely into place, 18m down. The camera had to be able to withstand the radiation inside the reactor and cope with the temperature — it was cooled with nitrogen.

He added: ‘It was particularly difficult when we got below the heating coils. There was nothing to keep the metal hot, so we had to get it all out before it solidified.’ The camera/pumphead had its own heating system, to prevent it getting stuck in solidifying sodium.

Removing the 900 tonnes of primary coolant and 600 tonnes of secondary coolant (which transferred heat from the reactor to the steam-raising system) has left the plant’s interior coated with a sodium residue which, said Husband, has the texture of porridge.

To remove it, the plant will be treated with a water vapour/nitrogen (WVN) process using a low concentration, between 4 per cent and 15 per cent, of water vapour in a nitrogen atmosphere to convert the residue into sodium hydroxide, which can later be dissolved and pumped out of the reactor. A similar NaK destruction plant is in the commissioning stages at DFR, which will also be treated by WVN. Both reactors will then be dismantled using remote cutting and handling equipment, and the parts steam-cleaned and sealed into cement-filled boxes for long-term storage.

Another tricky part of the site is the shaft. Originally dug to remove rock spoil from a conduit for discharging treated water, the 65m deep, 4.6m diameter shaft was used between 1958 and 1977 to dump intermediate level waste. Some of this contained sodium and, because water seeped into the shaft, in 1977 there was a hydrogen explosion in the gases above the waste. That ended its use as a dump.

Now, all 800m3 of waste has to be removed, classified, and treated for disposal or storage. The first phase of this project was to prevent water seeping into the shaft and becoming contaminated. To do this, John Whitfield led a team that drilled 200 boreholes around the shaft and pumped in a grouting mixture containing particles less than 20µm across.

‘The local rock has fissures only 25-100µm wide,’ said Whitfield. ‘We needed to squeeze the grout into the fissures at about 50 bar.’

The process, now complete, has cut water volumes flowing into the shaft from 15m3 to 1-1.5m3/day, he said. A test line of machinery is being assembled to extract, shred, clean and encapsulate the drums, metal components and other waste inside the shaft.

it will take until the late 2070s to remove all the waste from the site and the end-point of the project, budgeted at almost £3bn, is scheduled for 2300.

By then, DSRL hopes the only sign the plant was ever there will be the steel Golf Ball — a monument to a technology whose time never came.