The nuclear industry hopes that small modular reactors (SMRs) will become familiar workhorses of the energy system; supplied as a package, delivered in modules that can be quickly and easily assembled on-site and providing heat or power to neighbouring customers or to the grid as a whole.

In theory, using a small modular technology that is quickly rolled out in numbers allows the supply chain to make efficiencies, builds operating experience for owners and investors alike and drives the technology down the “cost curve”. That process has become familiar in the energy industry as wind turbines, solar PV and batteries have been rolled out in numbers that have risen from a few hundred units to become a mainstay of the energy industry in less than two decades. In fact, the Intergovernmental Panel on Climate Change (IPCC) noted in its March report that between 2010 and 2019 “sustained decreases” saw unit costs fall by 85% for solar PV, by 55% for wind power and by 85% for lithium-ion batteries. That came with deployment at the end of the period 10 times that at the start for solar PV.

Can nuclear SMRs follow the same route? Or has the industry’s current scattergun approach limited its opportunity to follow other clean energy technologies down the cost curve?

There are several consistent aspects to technologies that have successfully negotiated the curve. They are (initially) small; a single familiar technology has been used; it has been rolled out in numbers; they can be accommodated on a variety of sites; experience of the first projects has been quickly absorbed; and that data has given third party financiers confidence to invest in follow-on projects. Lithium-ion batteries, for example, are being rolled out at gigawatt scale, yet the largest installations, housed in facilities very like shipping containers, contain arrays of familiar small batteries. Innovation has been required in operating them as a mass array, integrating them with the grid, developing markets for the services they can offer (response, rather than baseload power) and modelling and managing the best operating regimes. Now there is ‘market pull’ as electricity system operators value the services they can provide.

SMRs are still at the ‘market push’ stage. The International Nuclear Energy Agency (IAEA) says it is aware of more than 80 designs that have been mooted, but its Advanced Reactors Information System (ARIS) lists 48 reactor designs in the SMR category that are at some point in the design phase. In fact, nearly a dozen of those actually fall outside IAEA’s own SMR definition, which puts an upper limit of 300 MWe on the reactor rating and the 300 MWe designation also means some of the more elderly conventional designs, such as India’s PHWR, are included in IAEA SMR listings.

The IAEA’s ARIS list has seven distinct types of technology at an early or late stage of design, with designs varying within each class. A dozen are versions of light water reactors that might be considered closest to the PWRs and BWRs that currently form the majority of working reactors. Seven are based around lead coolant, nine around sodium coolant and five around lead-bismuth coolant. Six have helium coolant and five are based around fluoride and graphite. Two combine a graphite moderator with a molten salt coolant.

Reducing costs by replication requires ‘learning by doing’ – ie building the reactors. That requires several issues to fall into place: a well-understood design, ideally with an operating track record, and a pipeline of sites that gives the supply chain confidence to invest. Where has the SMR industry reached in that process?


An SMR barrier that is not faced by many other types of electricity generation is the nuclear licence. This requires both a licence for the technology and for the site and it is a bottleneck that each technology has to pass through. The first to do so was NuScale, which won design approval for the six-module VOYGR variant from the Nuclear Regulatory Commission in January this year. However, design approval for the technology does not necessarily mean it can be installed at a chosen site. The US NRC approval for NuScale allows a utility or other operator to reference the design, but the operator has still has to apply for a combined licence to build and operate it as a nuclear power plant anywhere in the USA.

Likely to be the next to receive design approval is GE Hitachi’s BWRX-300, which “leverages the design and licensing basis” of GEH’s US NRC-certified ESBWR. It was entered into the US NRC licensing process in 2020. BWRX-300 has already completed phases one and two of the Canadian Nuclear Safety Commission’s (CNSC) vendor design review process and in December 2022 GEH announced that it has submitted it to the UK’s Generic Design Assessment (GDA).

The UK’s Office for Nuclear Regulation (ONR) also last year took on GDA of a new reactor from the UK’s Rolls Royce SMR. The reactor is described as an SMR by the company, although at 440 MWe it is considerably above the IAEA’s upper limit to meet that definition.

All three of these designs are based on well-known nuclear technology that has been deployed and has a track record of operation. Although the SMR version has numerous differences, these designs are likely to have the familiarity needed for mass rollout.

In the lead

What has been the progress towards choosing sites for the potential SMR fleet?

In contrast to large nuclear units, which are necessarily much more influenced by site characteristics, the key requirement for SMRs is replication. That means the site should be able to host several units or there should be a large number of potential sites available. It is no surprise that the technologies that have passed furthest through the design licensing process have a head start when naming sites for deployment.

NuScale’s progress through the licensing process has given it a strong position in raising potential buyer interest. It is selling its 77 MWe NuScale plant in three, six or 12-unit versions branded as ‘VOYGR’ and it is working with Utah Associated Municipal Power Systems (UAMPS) on a combined licence application for a six-module plant to be sited at the Idaho National Laboratory.

The application will be submitted to the US Nuclear Regulatory Commission (NRC) next year and is targeting units to start up in 2029 and 2030. The UAMPS consortium recently reapproved the project despite cost increases. UAMPS said a new budget plan “will move the small modular nuclear reactor project into an aggressive 2023 work plan, which focuses on completing the preparation of the application to construct and operate the plant, to be submitted to the Nuclear Regulatory Commission in January 2024.” This year it will also procure long lead material – NuScale Power has already placed an upper reactor pressure vessel order with Doosan Enerbility – and will develop a more detailed construction estimate.

NuScale has signed agreements with two other utilities that could open the door to more deployment. They are via Dairyland Power (which serves Wisconsin, Minnesota, Iowa and Illinois) and Associated Electric Cooperative (which serves 2 million people in Missouri).

In addition, NuScale Power has also unveiled a large number of Memoranda of Understanding with potential site owners in other countries. The most advanced is Romania’s RoPower Nuclear, which is owned equally by S.N. Nuclearelectrica and Nova Power & Gas, and which signed a contract with NuScale Power for front-end engineering and design (FEED) work in December. Phase 1 of the FEED work awarded to NuScale will define the major site and specific inputs for a VOYGR-6 plant to be deployed at the Doicesti power station. Importantly, Doicesti was not previously a nuclear site and instead had small coal-fired generation units.

Similarly, in Poland, NuScale now has agreements with companies in which it is providing ‘support’ as they investigate deploying VOYGR at sites that have not hosted nuclear technology in the past. One is a joint venture of Unimot (which offers Polish wholesale and retail customers fuel products, gas and electricity) and Getka (a US energy company) and the other is KGHM, a Polish metals extraction company and a large industrial energy user. KGHM has already begun discussions with the National Atomic Energy Agency (NAEA) and has signed an early works agreement with NuScale.

NuScale can also point to Memoranda of Understanding in Europe with Bulgaria, Ukraine and Czechia. Outside Europe, it has similar agreements in Jordan and Kazakhstan. All, of course, require the reactor to be licensed.

Also securing potential customers is GE Hitachi Nuclear Energy’s BWRX-300 SMR. It secured its first site last year when Ontario Power Generation selected it for deployment at Darlington. OPG aims to complete the first commercial construction at the site in 2028. Notably, GEH has signed project delivery agreements with SNC-Lavalin and Aecon for that project. Design acceptance by CNSC will give a head start for a second deployment by SaskPower which has selected the BWRX-300 for potential deployment in Saskatchewan, although that is timed for the mid-2030s.

This year GEH also announced that the Tennessee Valley Authority (TVA) has begun planning and preliminary licensing for the potential deployment of a BWRX-300 at its Clinch River site in Tennessee.

The regulatory approval process has been kicked off in Poland, where Orlen Synthos Green Energy (OSGE) and its partners have submitted an application to Poland’s National Atomic Energy Agency for assessment of the BWRX-300. OSGE says it wants to deploy “at least 10” BWRX-300s with the first in operation by the end of this decade. In addition, Fermi Energia announced that it had selected the BWRX-300 for potential deployment in Estonia.

Rolls Royce has not yet secured a site in its home market for its SMR. It believes the reactor will be of interest to so-called ‘industrial clusters’, areas which the UK has given a particular focus for decarbonisation. Government funding for the industrial clusters has been directed towards the development of carbon capture and storage (CCS) or hydrogen production, rather than opening the sites for new nuclear. The new nuclear option is currently restricted to specific sites in England and Wales (Scotland has a moratorium on new nuclear) identified when the government decided to restart nuclear build 20 years ago. In the end, the sites identified were previous nuclear sites. Planning law reform currently underway should widen that pool of sites but, currently, industrial sites would require special permission from the government. However, Rolls Royce has also found willing partners in Poland. The state-owned Industria business, a subsidiary of the Industrial Development Agency JSC, has selected the Rolls-Royce SMR for the Central Hydrogen Cluster, with plans to produce 50,000 tonnes of low-carbon hydrogen each year. It wants up to three units and Rolls-Royce SMR said that may open the door to “opportunities to replace more than 8 GW of coal-fired power plants in southern Poland with SMRs throughout the 2030s”.


NuScale, GEH and Rolls Royce have sought design approval first and built customer numbers as the reactor advances through the process. But X-Energy may undertake the process in partnership.

X-Energy is part of the Department of Energy’s (DOE) Advanced Reactor Demonstration Program (ARDP) and will receive $1.2bn in cost-shared funding from DOE over the life of the ARDP effort. Now it has signed a joint development agreement (JDA) with US chemicals conglomerate Dow to use X-Energy’s Xe-100 high-temperature gas-cooled reactor for process heat and electricity. The JDA scope includes the preparation and submission of a Construction Permit application to the U.S. Nuclear Regulatory Commission (NRC). X-energy is said to be preparing to submit a license application to the NRC.

XE-100 Reactor. (Credit: X Energy, LLC)

They plan to build X-Energy’s first unit at a Dow industrial site – one of several the company has along the Gulf Coast in Louisiana and Texas. A site is expected to be chosen by the end of this year.

Dow represents a fleet deployment opportunity in itself – it operates manufacturing sites in 31 countries where the Xe-100 could be used to decarbonise operations. It may join forces with X-Energy to license the technology to other industrial customers and power utilities. X-Energy had previously been in discussion with Energy Northwest and Grant County in Washington state, which would now have follow-on units.

Are we there yet?

Clearly, the Polish market will be a kick-starter for the industry. That presents its own risks: the European market for new and existing nuclear is made more problematic because some of Poland’s fellow EU members are opposed to any type of nuclear power and will object to it. Equally, however, EU members would like to see Poland able to close its coal industry and clean up its energy industry as a whole and that may benefit SMRs.

What about the large number of other technologies competing to roll out? It is likely that their opportunity will come after mass deployment of the leading options has created market ‘pull’. If the model of a small-scale reactor that is used on an industrial site is familiar to investors and to the public, it lowers those risks to deployment and can allow new technology to be presented to an investor comfortable with accepting risk.

Again, the battery market is instructive. Mass deployment of lithium-ion batteries has developed both the supply chain for delivery and installation and the skill set for making returns for investors. Now the ‘market pull’ is there, attention has turned to the next technology. A recent analysis from Bloomberg New Energy Finance says “Sodium ion batteries are more expensive than lithium-ion today because of low volumes and underdeveloped supply chains. But BNEF sees potential for material savings and energy-density improvements that would provide a viable pathway for sodium ion cells to cost half what lithium iron phosphate does today”.

The Linglong 1 SMR in China. (Credit: China National Nuclear Corporation)

The nuclear industry is grappling with deploying SMRs but two or three versions are showing the characteristics that have been successful for other asset classes. The technology is relatively familiar, while the small size of units means they will be deploying in 10s, rather than singly. The question will be whether that can step up to the hundreds or thousands, to gain the true benefits of replication. That may see promising designs fall by the wayside – but that is an experience familiar across asset classes, when falling costs due to replication overtake a technology that would, on paper, look more economic. In that case, experience suggests the broader energy industry and the investment community will “not let the best be the enemy of the good” and take a pragmatic decision: familiarity wins.

This article first appeared in Nuclear Engineering International magazine.