Dounreay destined for demolition: endgame for an atomic icon
Image credit: NDA/DSRL
Preparations are under way to demolish the landmark Dounreay sphere on the windswept north coast of Scotland. The complexity of the project is a lesson on the need for foresight in new energy developments.
The iconic sphere nestled among a squat series of buildings at Dounreay marks the location of a pioneering site in 20th century nuclear technology. This part of the north coast of Caithness, Scotland, was the home of the UK’s fast-breeder reactor programme. It was created primarily to make electricity and plutonium, but was also used for reprocessing nuclear material and carrying out cutting-edge scientific and engineering experiments.
The sphere - inevitably dubbed the ‘golf ball’ in this, the ancestral home of the good walk spoiled - housed the Dounreay Fast Reactor (DFR), hailed for its pioneering position in global nuclear engineering. In 1962, it was the first fast reactor in the world to supply electricity to the national grid. Now, however, it is one of the most challenging nuclear sites in the world to decommission.
The pioneering efforts of the atomic scientists and their willingness to push boundaries has left a legacy of complicated, hazardous decommissioning work. Dismantling radioactive machinery is fraught with difficulty, and today’s engineers must negotiate ageing infrastructure in inaccessible, hazardous areas.
Old nuclear sites were built perhaps with a dim assumption that we genuiuses of the future would have figured out how to demolish them. But simple demolition is not enough at Dounreay. Many parts of the site hold dangerous nuclear material and levels of radiation that mean workers must be protected. Today’s security climate also means that restrictions are in place on moving nuclear material around.
The design, use and piecemeal adaptation of old nuclear sites like Dounreay have given today’s engineers complex problems to overcome, and a keen understanding of the need to think ahead. The use of proven technology is encouraged, as well as simple processes wherever possible.
Dounreay has a chequered history when it comes to radioactive waste management but the site is now a popular venue for international engineers – among them recently a Japanese delegation – to learn how to dismantle specialised atomic sites.
The nature of the experiments that took place in DFR meant that when it was shut down in 1977, some of its fuel was stuck in the reactor and remains there today. Dealing with the reactor’s volatile liquid metal coolant was also a big challenge, which has since been overcome, making DFR less hazardous for those dealing with the reactor contents.
The heart of the DFR sphere is the reactor, now switched off but still very radioactive. Before the reactor can be dismantled and the outer sphere shell taken down, workers must figure out how to remove hundreds of elements that have remained in place since the shutdown.
The reactor is essentially a large, metal bucket set into floor level inside the sphere. The bucket contains a very long honeycomb, where hundreds of smaller channels held metal rods with an outer metal cladding. The rods had different purposes in the reactor, including to make plutonium, reflectors to bounce energy back into the nuclear chain reaction, and fuel rods.
Senior project manager Ron Hibbert is responsible for the complex engineering work required to empty the reactor. He explains that the work is aligned with the Nuclear Decommissioning Authority’s Magnox Operating Programme (MOP), which sets a timeline for when the reactor should be emptied and its nuclear load transported to Sellafield in England.
Even getting to this point of removing the fuel has been a very long journey – the removal of the reactor’s liquid metal coolant, composed of sodium and potassium, was a significant challenge given its hazardous, volatile nature and its location in a tangled knot of complex pipework. Some of it had solidified and a special chemical reaction process had to be developed and tested before work could begin on the real thing.
The decommissioning of DFR has involved demolishing buildings and putting up new ones, as well as installing safer, more reliable services. Even so, engineers often encounter new problems as work proceeds.
Issues emerged in the Dounreay golf ball around the complicated sequence of operations needed to remove the fuel. The original circular metal rail from the 1950s, upon which the huge 25-tonne DFR crane runs, had moved out of position, requiring a new approach to moving the crane.
Despite this and other issues, Hibbert remains optimistic that the project can meet the timeline required by the MOP.
He says: “On the breeder project we’d had a really good year but we then had some problems with previously installed equipment. We had a number of difficulties trying to set to work with previous infrastructure.”
The knock-on effect is that there is now a plan to rotate the polar crane through a different angle in order to access the reactor elements and move the sphere’s special element transfer flask from that point. Issues have also come to light with a piece of equipment where cables became tangled during various configurations needed to access the breeder elements through the reactor’s rotating shield. A new tool has been devised to tackle the problem.
Hibbert comments that the aim for the new replacement was to keep to simple engineering principles and produce a locally built tool rather than design a complicated, expensive solution.
“We have replaced the mast with a very simple retrieval system. One of the things we are most proud of is that everything you see mounted on the reactor top apart from the retrieval cell has been designed by the project team and supplied by local contractors.
“We are very proud of the fact that there is local capacity to do this; there are highly skilled local companies and knowledge within the site that helps us work out these problems.”
Radiation and contamination always complicate dismantling work as there are risks to staff and the creation of more nuclear waste to consider.
“What I learned, and it stuck with me, is that if you are doing work in the reactor you have to be very confident of what you are doing before you introduce equipment because there will be very little chance to modify things later,” says Hibbert.
“As soon as you do anything in the reactor, you are potentially contaminating things and there is no chance of getting the equipment back off site. Modifying on-site is difficult because you then need a radiation-controlled area, welding restrictions and so on.
“Your best shot is to test before you put it in. So I have used that approach, which I learned on various projects at the Dounreay Prototype Fast Reactor. We set up a trial facility at the T3UK site. What it did was replicate the top function of the reactor, made sure the heights were right and had a simulated section of the core area.”
T3UK is an engineering test facility and educational campus in the north of Scotland aimed at the nuclear, marine and energy markets. The test structure was a mock-up of a section of the DFR reactor with a test rig above it.
He continues: “We developed our replacement design, built a trial version of it, and we evolved that. It was needed to give us the confidence that where we were going with the design was right. We needed to be able to reach everything while it functioned as intended.
“We kept that at T3UK for various reasons including as a ‘hot standby’ so that if there is a problem we can quickly study it offsite, come up with a solution and then replicate that in the actual reactor.
“There is damaged fuel in the reactor so we will retain the rig there for the further tooling and equipment needed. When we take out the 500 or 600 easier elements, that gives us time to develop solutions for the challenging ones – that’s the philosophy.”
Using the new design is part of a complex 3D jigsaw where some parts, such as the crane and reactor shields, rotate in a circle and others, such as the element transfer container, need to move to the correct position horizontally and vertically to take elements from the core to a special pipe where they are sent downwards for inspection. After this, they reach the building where they are cut up in shielded cells for packaging and transport.
Hibbert continues: “The new design has worked very well for us. That was a challenge in itself with the restrictions on the dimensions, and it has to work in two halves; it’s like one shell inside another. But at the same time, half that diameter is still a significant size, it is a bit heavy and fragile. The transport frame had to come in here and go into position so that we could, safely and in a controlled manner, lift it into there after a careful assembly procedure.
“You can say it in a sentence or two but the actual work it took to do that was significant. At the time it was a contaminated area, so the last thing we wanted was to have people spending longer there than they needed to. That is why we did all the work off-site, so that what you have to do is come in, set it in place and bolt it down. That saved a huge amount of radiological exposure.”
‘A lot of effort has been put into inactive commissioning of this system. If anything breaks, we have demonstrated that we can replace or repair it remotely.’
To make the jigsaw even more complicated, some of the rods are stuck in the reactor channels. A ‘top plate’ metal structure in the reactor has been cut to free up some elements, and more needs to be cut to access headless rods that are shorter than the others. Different grabs are needed for the new tool.
Some of the most damaged rods were affected by overheating in the reactor during experiments decades ago. In some instances, the reactor coolant was supposed to keep temperatures under control but, in places, the elements got too hot for them to stay in the right place and shape. Late in the life of the reactor, materials were deliberately subjected to extreme conditions, including coolant boiling. Localised parts of the reactor may have reached up to 900°C, which has caused fuel failures.
Preliminary work has shown that some elements are not in a bad condition, so the current aim is to focus on taking these out, while work in the neighbouring building where elements are cut up and packaged for transport is under way.
Hibbert says: “We’re initially using a simple philosophy to extract them. In reality, we think that between 500 and 600 elements are straightforward to remove. What we found when we worked in the reactor last year is that the elements further from the centre should be relatively straightforward to take out. Because the sphere work is quicker than what we have to do in the breeder building, there is an opportunity to do other work between times.
“To underpin our assumptions, we went in and checked this. This tested our assumption and proved some elements are not stuck at the bottom. The elements might not come right out to the top when you pull them, but all the indications are that the initial survey was right.”
Additional work has also been needed to move the project forward, in the form of using the T3UK engineering facility to look at why a transfer flask wasn’t working as expected. By using a deep pit in the test centre, workers were able to see that a grab was twisting inside the flask when it should not have been.
“You can’t legislate for ageing infrastructure,” says Hibbert. “What assumptions can you make? What’s going to be good? What’s going to be bad? We go through a thorough maintenance regime but these types of challenges happen and, to be honest, there are other people going through similar issues at other sites.”
Sharing experiences with other sites has shown the team that engineering problems are faced at many industrial locations that require simple, problem-solving approaches.
After travelling down a pipe and being inspected, the elements are stripped of their cladding and put into shielded packaging. By then they should be the right length, width, weight and composition for the heavily shielded flask, which is used for transport off site.
Project manager Chris Wratten explains that the work is carried out in shielded concrete cells with thick lead glass windows and a nitrogen atmosphere for safety reasons. Remote grabbing arms and a camera system are used to carry out the work. One cell is where small amounts of liquid metal coolant are still present on the radioactive metal pieces; an adjacent cell is used after it has been cleaned and is being packaged for transport. The DFR material is destined for the reprocessing plant at Sellafield.
“The long-term plan for the cladding is that it goes into the intermediate-level waste store and it remains there until the radioactivity levels go down. All the waste stays on site and all the fuel goes to Sellafield,” Wratten explains.
He says the current approach means that in newer buildings, the services such as the electrical supply can all be accessed and maintained easily away from the radioactive areas. The end result is that when this newer plant needs to be decommissioned, it will be easier to take apart and there will be less nuclear waste to deal with.
Wratten’s approach echoes that taken in the sphere work, of testing equipment first before any radioactive material is involved.
“There has been a lot of effort put into inactive commissioning of this system so that we don’t have to then go back in and make any changes. With every piece of kit in the cell, we have demonstrated that if it breaks, we can replace it or repair it remotely. We know it’s robust and maintainable. We don’t then have to put people into a hazardous area. Something that is easy to do by hand is not easy with the remote grabs. It’s about making things simple.”
For older shielded cells on site, a system of periscopes and mirrors was used to view the nuclear materials being worked with, as well as special windows. Today, standard cameras are used to get a better view than the slightly distorted 2D view through the lead glass window.
“We use standard, cheap, off-the-shelf cameras. You could specify expensive, radiation-tolerant cameras but they’re not as good and don’t have the same functionality as the normal cameras.
“Because the radiation doses aren’t that high in this cell, we can use standard cameras from other industries and they don’t have the added cost of being radiation-tolerant. If they do fail, they’re easily replaceable. Even then, we haven’t actually had one fail yet in the lifetime of the plant, and you can see fine detail from them.”
Removing the elements from the ‘golf ball’ sphere is one of the last jobs to be done at DFR before the reactor is taken apart and the metal shell taken down. Last year, the site operators submitted a planning application to the local council for a series of projects from 2018 towards the site shutdown. These jobs include some major changes to the Dounreay skyline, which will be cleared of its Cold War icon.