The case for small modular reactors (SMRs) is compelling. Unlike traditional nuclear power plants that take years – or even decades – to build, SMRs will largely be made inside factories. Identical reactors will roll off the production line, ready to be shipped to site, fixed together and switched on. Once the production line is up and running, the companies selling SMRs will – in theory – be able to predictably churn out multiple units a year.
While SMRs will generate far less power than traditional reactors, they’ll be cheaper and quicker to make. That should make them easier to finance and mean investors get their money back sooner, all while delivering carbon-free power faster.
Several countries are in the race to produce SMRs. In the UK, Rolls-Royce’s SMR design has been selected for use on a site in Anglesey, and a contract to start work on three of the reactors was signed in April this year.
However, until now, no one has actually built any kind of nuclear reactor in a factory environment. Building one is typically a highly complex construction project. A company might make a few versions of the same design, but each facility is pretty much unique, custom-built on site.
Perhaps the closest analogy would be modular housing, in which whole buildings – including their walls, roofs and rooms – are pieced together inside a workshop before being shipped to a plot of land. This example demonstrates how structures traditionally built on a construction site could largely be transferred to factory processes, yet the scale and complexity of nuclear facilities is hardly the same as a prefab house.
So what might actually be made inside SMR factories? And, more generally, how might this emerging industry work?
Early days
“We are in the developing stages,” says George Borovas, a lawyer and nuclear specialist at law firm Hunton, who advises companies on the financing of SMR projects. At present, SMR firms are still putting together plans. At most, he says, some might be placing orders for certain “long-lead items”, which they may need to make reactors in the future.
Nevertheless, this is an industry that appears to be burgeoning. The World Nuclear Association has a live map of SMR projects around the globe on its website. A handful have already been built, while close to 100 projects are at various stages of (mainly early) development.
To turn all these plans into power-generating SMRs, factories will need to be built to start making the reactors themselves – not to mention all the other components and structures that will come together to form an SMR.
Not just one factory
The promise of SMRs is that they will allow companies to build multiple reactors in factory conditions. However, it’s unlikely we will see one big plant turning out complete products ready to ship. Instead, the nuclear company – be that Rolls-Royce, Westinghouse, Holtec International or anyone else – will take on a coordinating role, working with their supply chains, who will produce large portions of the final SMR.
That being said, nuclear companies will probably make some of the most important or secret elements themselves in-house. “A lot of [these companies] are recognising they have parts of their technology that are bespoke to them and they want to keep it that way,” says Charles Carpenter, head of research for the University of Sheffield Advanced Manufacturing Research Centre’s nuclear manufacturing group.
Laser welding at the University of Sheffield AMRC
Rather than outsourcing the production of their intellectual property, companies will likely make these elements themselves in-house. For some reactor companies, this unique aspect might be a coolant or the fuels, for example.
These companies will also certify kit produced by their suppliers and then assemble it inside their own factories. Carpenter continues: “You’ll see these modules coming offline with mechanical, electrical and pump piping modules. You’ll see modules with valves in them, with tanks, with pumps in them, with electrical units. Those are going to be built in factory. They’ll be tested in the factories themselves and certified there.”
Built to order
One example of this would be US nuclear firm Holtec International’s plans to build their small modular reactor, the SMR-300. Patrick O’Brien, the company’s head of communications, says they expect to build “steam generators, reactor pressure vessels, modules for the construction of containment structures and other ancillary components” at its factory in Camden, New Jersey.
Making these components will require some advanced manufacturing technologies. O’Brien lists equipment for “welding, rolling, machining, including the largest plate roller in North America, which can roll 7in cold steel and 10in heated steel”.
The first SMRs will likely be made using traditional manufacturing methods, Carpenter reckons. However, once the process starts to scale up, we’re likely to see more advanced methods such as electron beam welding or laser welding, which will allow workers to be far more productive.
Layers upon layers
“There’s not just a car factory, like one building that basically does it all,” explains Adam Locke, nuclear sector lead at construction firm Laing O’Rourke. Locke points out that auto manufacturers don’t build all the parts that go into a car themselves; instead, they assemble various widgets made by suppliers elsewhere. “You’ve got layers of factory. There’s a factory somewhere that makes dashboards, another one that makes electronics. They arrive at the car factory and they’re quality-controlled and certified” before being fixed together.
This is how SMRs are likely to be built in the future. Dense supply chains will feed in a whole range of elements for the final SMR, including pipes, pumps, valves and thousands of other components.
However, unlike a car factory, where finished vehicles roll out the gate, a large part of the final fabrication can only ever be done on site. Indeed, it’s likely that much of the final ‘product’ will have never passed through a factory belonging to an SMR company.
Holtec plans to build components at its factory in New Jersey
For example, Laing O’Rourke makes modular concrete structures that could be used in the build of a final SMR. It wouldn’t make sense to ship these to the SMR company’s factory – instead they’d just be shipped to the site and fixed together around the reactor itself.
Building an SMR will, in a sense, be like putting together an extremely complex Lego set, suggests Locke. While the reactor itself may arrive at the site ready-made, it will also need to be connected up to a dizzying array of components, pipes, valves and a concrete structure in a predefined way.
Even though SMRs are an order of magnitude smaller than traditional nuclear power plants, there will probably still be tens of thousands of components that need to be attached to one another. Planning out the order that things get fixed together will be fiendishly complex. This will, however, be made somewhat easier thanks to digitalisation. Having some kind of digital twin will allow builders on site “to just check on a tablet” to work out how to attach pieces together, believes Carpenter. As long as everything’s been made and delivered according to plan, the build process should be relatively straightforward.
Supply chain reaction
Manufacturing SMRs will, clearly, be far from a simple process. And there are a number of significant challenges companies hoping to build these products will face.
One difficulty will be around setting up the supply chain. Borovas says: “The supply chain needs purchase orders, right? Before they say, ‘OK, I’m going to start manufacturing this component’, they want to sign a contract that says, ‘I’m going to pay you a certain down payment and then milestone payments’.”
This creates a chicken-and-egg situation. If, for instance, a manufacturer is to start building certain valves for an SMR customer, the SMR firm needs to provide some funding up front. But since SMR firms aren’t yet making money – or may not even have a final design – they might not be ready to provide that upfront cash. Until definite orders are agreed, it will be challenging to get the supply chain to start gearing up.
Start-up challenges
Another issue is cost. SMRs should, in principle, be cheaper to make than traditional gigawatt-size reactors, but they’re still going to be pricey.
Borovas explains: “In nuclear, just like in every business in the world, when you’re still dealing with a first of a kind, you’re always going to have that challenge with respect to timing into budget. So, yes, [SMRs are] going to be cheaper, but the first one is probably going to be more expensive than you’d think.”
The issues involved in building the first of a kind are particularly challenging when it comes to nuclear reactors. Carpenter explains that, with traditional reactors, having a one-off design means engineers can tweak and customise as they go. But creating a design for mass-manufacture poses its own set of challenges.
Essentially, your initial design has to be your final design. If you’re setting up large manufacturing facilities to mass-produce multiple copies of the same reactor, there is very little room for error with the first product. Making any changes to the design would be very costly if you’ve already set up a factory to work in a certain way.
But just as significant is regulatory approval. Any nuclear reactor design must go through very rigorous checks to ensure safety and compliance. So, even the smallest changes between the first design and subsequent models would require a manufacturer to go through the entire approval process again. This could slow things down, and any delays could shake customers’ confidence in the basic premise of SMRs.
Interface issues
As we’ve seen, a fully functioning SMR isn’t going to leave the factory and get plonked down on a site. Rather, multiple parts – from the reactor itself to all the pipes, valves and turbines, through to the concrete shells and beyond – will arrive at a site somewhere to be connected. This raises issues about interfacing: will all the parts actually connect as expected?
Carpenter says: “All your piping connections, all your electrical connections need to connect, fit, need to be able to come together… you don’t want to be bending pipes too much on site or moving them around.”
Again, this is because any significant change to the design of a nuclear facility needs regulatory approval. If builders at the site have to make lots of changes to the plan to get different components to work together, this could introduce real risks when dealing with potentially hazardous plants.
Rolls-Royce's SMR design has been selected for use on a site in Anglesey
A final big challenge for SMRs is the workforce required to make these things. The problem, according to Carpenter, is that people with the required know-how are in short supply and high demand.
Many of the skills needed to manufacture SMRs are also needed in several other energy-related industries, from wind to solar to hydrogen, as well as oil and gas. People who know how to weld or fit pipes are also in demand in defence industries. Countries will need “national strategies on how we’re going to get our workforce and skills and supply chain ready for a rapid increase of the need” for these skills, Carpenter believes.
Timing is everything
None of the challenges of setting up SMR factories or their networks of suppliers are insurmountable. And, with many governments expressing renewed interest in nuclear energy, the stars might just align for these facilities to take off.
Locke points to the growing demand for power from data centres, the need for carbon-free energy to meet net-zero targets and countries’ desire for energy independence following recent fuel price shocks.
“You might say I’m biased, but I do think now’s the time for this [emerging industry] to take shape because, basically, it needs to,” he says.
Extracted from IMechE website, read more here