Introduction

Nuclear energy by fission is a proven greenhouse-gas-free energy source1 that has struggled to achieve public acceptance and to keep costs of construction competitive. Indeed, the lack of wider adoption of nuclear energy is due more to public concerns and economic risk than to any particular technological limitation. However, with the growing impact of climate change and the ever-growing need for more abundant energy, this balance is shifting and there is renewed focus on the development of fission energy systems. This new era is perhaps unique with the relatively large investment of private industry in novel reactor concepts. In reality, many of these ideas are not new, but are seeing instead a conceptual rebirth due to the desire to build even more economical, efficient, and passively safe reactors. All of this said, and despite the overall success of the existing nuclear fleet in generating clean and consistent energy, there remain materials challenges that, once overcome, can further enhance the performance of these reactor systems. This includes the heart of the reactor, the nuclear fuel.

There are a number of reactor designs, both commercial and for research, that leverage different behaviors to extract the greatest amount of energy from the fuel in a safe and economical manner. This means, correspondingly, that the fuel form varies with reactor type. For example, water-cooled reactors utilize uranium oxide (UO2), fast reactors with molten metal coolant use metallic fuel or ceramics such as uranium nitride (UN) or mixed oxide (MOX), gas reactors often rely on TRi-structural ISOtropic (TRISO) particle fuel or uranium carbide, and molten salt reactors take advantage of the benefits associated with having the fuel in a liquid rather than the solid form used in other designs. Many basic concepts were conceived some time ago. However, this should not be mistaken for lack of development or progress, as exemplified by today’s UO2 fuel for light water reactors running at higher power and having a much higher reliability than in the early days. A similar development is envisioned for advanced reactors and fuels once they have been deployed. Further, there is a significant opportunity to reduce the cost and time needed to deploy fuel for both current and advanced reactors.2,3

The current state of nuclear fuels

Since the inception and first demonstration of a sustained nuclear reaction, Chicago Pile-1,4 there have been a variety of nuclear fuel forms that have been pursued with the ultimate goals of efficiency and reliability. Some of the most prominent fuel types, along with their relative properties, are summarized in Figure 1. When considering the properties of a nuclear fuel, a number of factors rise to prominence. These include fissile density—the potential amount of energy stored in the fuel, typically in the form of 235U and Pu; ease of fabrication, as this directly relates to economics of the reactor; and thermomechanical properties, including the stability of the fuel and the ability to extract the energy out as it burns. The ability to accommodate fission products, particularly fission gases, is another important factor dictating the overall performance of a nuclear fuel. Other factors are also worth considering, such as the reprocessability of the material.

Figure 1
figure 1

A number of different fuel types have been proposed over the decades, each with its own relative strengths.

Full size image

Types of nuclear fuels

Despite the decades of research in and development of nuclear fuels, the predominant fuel used in reactors is still uranium dioxide—UO2 or urania.5 In fact, all reactors that generate electricity for the grid use oxide-based fuel, either as pellets, in TRISO form, or as (U,Pu)O2. Urania represents perhaps the best set of tradeoffs of performance versus limitations. It has a sufficiently high fissile density, exhibits good stability, and effective retention of fission products. In particular, UO2 is highly resistant to radiation damage. Further, UO2 is the most stable compound containing uranium. Certainly, irradiation leads to structural evolution of the material, but it is essentially impervious to amorphization, retaining its crystalline structure to high doses, which is advantageous for maintaining thermomechanical properties. Fuel swelling is lower than in many alternative ceramic and metallic fuel forms. However, it does suffer from relatively poor thermal conductivity, meaning that it is harder to extract energy from the material. Even though the fissile density is not a hindrance in today’s light water reactors, it contributes to limiting power and core cycle length. Further, the fissile density is countered either through the size of the reactor or increasing enrichment, which are important considerations for smaller reactors currently being developed. The application of UO2 in light water reactors (LWRs) and proposed small modular reactors6 of LWR design also benefits from compatibility with the water coolant, which means limited reaction rates if the cladding ruptures, and with the Zr cladding itself. For these reasons, most new reactor designs based on a water coolant use UO2 as the fuel.

MOX7 refers to mixed oxide nuclear fuel and consists of a uranium and plutonium solid solution (i.e., (U,Pu)O2). The benefits and drawbacks are similar to those of standard UO2; however, thermal conductivity is somewhat reduced. MOX provides neutronic benefits and can effectively reduce uranium enrichment requirements. It is also a good alternative for molten metal-cooled fast reactors. The interest in MOX fuel originated when fuel recycling and reprocessing was developed with an aim to close the nuclear fuel cycle. MOX is not actively considered in the United States, but it is being pursued in other countries for both light water and fast-reactor designs.

Other uranium-bearing ceramics have also been pursued with the goal of both increasing the fissile density and improving the thermal conductivity. For example, fuel forms based on borides,8 nitrides,9 carbides,8 and silicides10 have been developed and tested. Many of these structures, exhibiting much higher degrees of covalent bonding than the oxide, are consequently more prone to amorphization. Thus, there is a greater possibility that there is a complete structural phase transformation that would cause corresponding changes in relevant thermomechanical properties. However, for most power reactors, the operating temperatures are high enough to avoid amorphization during operation. For some chemistries, this may not be the case during startup or shutdown, for spent fuel, or for research reactors that operate at much lower temperatures. The nonoxide ceramic fuels typically exhibit higher swelling11 than UO2, in particular at high temperature and high burnups. Nevertheless, given appropriate understanding of these phenomena, designs and operational limits can be established to manage the risk. The nonoxide ceramics are appealing for all reactor types, but especially for those that do not have to worry about the potential for oxidation when exposed to the coolant. This is the case for fast reactors cooled by molten metals, gas-cooled reactors, and reactors cooled by heat-pipes. Specifically, uranium carbides have been explored in sodium-cooled and high-temperature gas-cooled reactors and uranium nitride is a candidate for lead-fast reactors. Uranium nitride and carbides are of interest for space nuclear power due to their thermal stability, fission density, and favorable thermal conductivity. Silicides and nitrides have also been investigated for use in water-cooled reactors, but require additional mitigation strategies to deal with oxidation in the event of a cladding breach.

TRISO12 is an interesting case. TRISO consists of a kernel of a uranium-bearing material, usually UCO but sometimes UO2, surrounded by various layers of carbides or carbon-bearing materials that both absorb and retain fission products and thus maintain structural integrity. Research on TRISO has been ongoing since its initial development in the 1960s but has seen new interest in recent years. Early designs relied on UO2 for the kernel, but to avoid internal pressure buildup due to the formation of CO gas, modern TRISO fuels have a composite kernel consisting of UO2 and UC2–x,13 although recently UN has been proposed for the kernel, which would bring the same benefits as UN in pellet form previously discussed.14 The main benefits of TRISO are its thermal stability and ability to retain fission products due to the coating layers. This has made it attractive for commercial reactor designs as it opens a clear licensing pathway. TRISO belongs to a broader class of particle fuels, which were proposed in gas-cooled pebble-bed reactors as far back as 194415 and is also the preferred choice in current pebble-bed designs. The TRISO particles are embedded in a compact that constitutes the pebbles. Recent designs have seen TRISO particles embedded in compacts using coolant strategies other than gas and/or pebble beds. One such example replaces the gas coolant with a molten salt as the cooling medium.16 Others embed the particles in a matrix and fuel rod type design.17

Metallic fuels,18 such as uranium alloys, are less common in commercial systems but are extensively used in more specialized situations, including military and space applications, and research reactors in the form of U–Mo, U–Zr, and U–Zr–H. However, they are a preferred option for sodium-cooled fast-reactor designs in the United States, which are being actively explored by US developers with established plans for a demonstration reactor. Metals represent the highest fissile density and thermal conductivity, but swell significantly, could amorphize, and exhibit chemical evolution leading to phase changes. There can also be detrimental interaction with cladding materials and they have a low melting point (though that is counteracted by the high thermal conductivity to some extent).

Molten salt fuels19 differ from the fuel types discussed thus far by being liquid rather than solid. This has the benefit of not having to worry about structural irradiation effects (i.e., there is no amorphization, swelling, or degradation in thermal conductivity). Heat transfer is also effective in the molten salt. The main challenge of molten salt reactors is corrosion of structural materials exposed to the salt. The base salt is either fluoride, for example, FliBe (LiF + BeF2) or FliNaK (LiF + NaF + KF), or chloride, for example, NaCl, KCl, MgCl2 or mixtures thereof, with uranium as the fuel element. Properties to be considered for the salt include density, melting points, viscosity, vapor pressures, and corrosion of piping.

Ongoing considerations for nuclear fuel development

Numerous nuclear fission reactor development efforts seek to enhance the ability to extract energy from the fission reaction safely and economically. Ultimately, objectives can be broadly categorized as either targeting enhanced materials performance and/or increasing the economic viability of the nuclear fuel. Both are closely tied to facilitating regulatory approval of the fuel.

Materials science

The primary challenge associated with nuclear fuels relates to the nonequilibrium evolution of these materials as they are burned in the reactor. That is, the fission process itself leads to structural and chemical evolution, meaning that the properties of the pristine material evolve.11 That evolution often involves the degradation of properties, and thus understanding it is critically important for being able to predict properties. Fission leads to high-energy particles that displace atoms, causing radiation damage. Depending on the material and the conditions, this can lead to the formation of defects that can aggregate, forming interstitial loops and voids, and in the limit of high dose swelling of the material. In other cases, the damage leads to structural changes such as amorphization, leading to even more drastic changes in properties. Thus, at a minimum, we want to be able to predict this evolution so that we can also determine how properties change with burnup, though the ultimate goal would be to identify fuel forms that are highly radiation-resistant. Fortunately, uranium dioxide is an example of a highly radiation-tolerant material, providing one reason for why it is so widely used. However, given its other limitations (such as low thermal conductivity), there are opportunities for improvement.

The fission process also leads to changes in the chemical composition of the fuel via the formation of fission products. These fission products could also change the thermomechanical properties of the fuel. In particular, fission gases such as xenon precipitate out in the form of gas bubbles and the release of the resulting pressure into the plenum of the fuel rod can result in catastropic overpressurization of the fuel rod. A primary challenge in developing new fuel forms is managing the inevitable fission product inventory, to retain it within the fuel in a way that has minimal impact on properties and limits swelling.

A key material property for nuclear fuel is thermal conductivity as it governs heat extraction. If the heat generated by fission cannot be extracted, the fuel will perform inefficiently. Uranium dioxide already has a relatively low thermal conductivity and radiation-induced changes only make it worse. Thus, a primary motivation for the development of new fuels is to enhance not only the initial thermal conductivity, but also retain higher levels of thermal conductivity as the fuel burns.

Finally, there are questions about the compatibility of the fuel with other reactor components. Specifically, the fuel may interact with the cladding and detrimentally impact properties such as melting temperature and mechanical properties.20 If the fuel comes into contact with the coolant, it is important to ensure that any reaction does not seriously impact fuel integrity. Reactivity on its own is not a problem, rather it comes down to the rate at which energy is produced. In practice, this means that the coolant must not react with the fuel in a very rapid and endothermic way. These aspects were two of the main drivers for the choice of UO2 plus zircaloy as the fuel and cladding in light water reactors.21

Economics

Economics is another key driver of fuel selection and design. If a fuel is costly to process or deploy, it will see little use. Again, uranium dioxide has advantages in ease of processing, but has a moderate fissile density. More efficient fuels can be developed if the fissile density can be increased. This is a driver in the development of other ceramic fuels in which the uranium density is higher.

If a given fuel type is expensive to make, it will also see less use, unless it has other significant advantages. TRISO is a much more complex fuel form, involving chemical deposition processes to create the various barrier layers. However, these are included precisely to control the release of fission products, keeping them within the fuel form and thus preventing them from interacting with other parts of the fuel pin or even reactor. Therefore, for certain systems, advantages of more complex fuel types could warrant consideration by reactor developers.

A key open question, at least in the United States and most of the nuclear-energy-producing nations in the world, is what to do with the spent fuel. Strategies have been developed, though few have been implemented, for the storage and/or reprocessing of spent nuclear fuel. Essentially, governments and utilities have deferred this question to the future. However, for the long-term sustainability of nuclear energy, the economics of waste treatment will become an ever-present factor and must be accounted for. Different fuel types will have different amenability to reprocessing. Reprocessing provides the option to use more of the fissile material and reduce the amount of long-lived fission product isotopes. An example is metal fuel, which can be reprocessed with relatively simple one-step electrorefining.

Ultimately, regardless of the fuel form, we would like to extract even more energy from the fuels, as this increases the overall efficiency and decreases the cost associated with the reactor. One reason that uranium dioxide is so prevalent is that collectively we have so much experience with the material—we know better how UO2 will evolve in a reactor than any other material. However, this does not mean we should be myopic to other fuel types and new reactor concepts that inherently account for the uncertainty in properties with burnup by, for example, utilizing advanced experimental and modeling techniques to develop and parameterize models that predict not only performance for a particular set of parameters, but also its distribution based on the uncertainties in the input going back to the fundamental atomic to microscale mechanisms that govern the response irradiation and temperature in the reactor.22

Safety and reactor design

The fuel form is one part of the reactor safety system. The mission is to prevent release of radioactive fission products to the environment and the public. The first barrier is the fuel itself. By containing fission products in the matrix, release is prevented. However, this is not possible to fully achieve as gaseous species, in particular, tend to exhibit at least fractional release, especially for high burnups and at high temperatures. If the fuel were to melt, more fission products would be released. All reactors have other barriers to prevent release, including cladding, containment buildings and, in the case of TRISO, the fuel particle itself (coating layers). Regardless of the reactor design, it is beneficial for the fuel to retain fission products, avoid fuel melting (unless it is a molten salt reactor), and avoid interaction with other containment systems such as the cladding. The latter is influenced by swelling, which is caused in part by the retention of the gas in the fuel pellet and could seem counter to preventing fission product release. However, it is not only the amount of gas retained but also its form (large versus small bubbles) that governs swelling. This all leads to the design of new reactors that can leverage the strengths and mitigate the weaknesses of various fuel forms. For example, the consequence of large swelling of metal fuels can largely be engineered away by starting with a sufficiently large gap between the fuel pellet and cladding.

History repeating itself

While there is an uptick in the activity surrounding the development of nuclear fuels, again motivated by new investments in advanced reactor designs, the reality is that most of the nuclear fuel forms being pursued are not new. In fact, as highlighted in Figure 2, most concepts and designs trace their origins to the first few decades after Fermi’s initial demonstration of a controlled nuclear reaction. Fermi’s pile, which demonstrated a sustained nuclear reaction in 1942, used natural uranium in a combination of both uranium metal and uranium dioxide as its fuel. After World War II, many alternative fuel types were explored with various levels of success and commercial deployment.

Figure 2
figure 2

Nuclear fuels research has been vigorous since the earliest days of Chicago Pile-1. Many new fuel concepts were developed and tested within the first few decades of that historic achievement. A number of them are seeing renewed interest for advanced reactor concepts. Photo credits: Row 1: University of Chicago Photographic Archive, apf2-00501, Hanna Holborn Gray Special Collections Research Center, University of Chicago Library. Wikipedia. Wikipedia. Los Alamos National Laboratory. Row 2: Photo courtesy of Argonne National Laboratory. Wikipedia. Department of Energy. Photo by Staff Photographer Seattle Post-Intelligencer, courtesy of MOHAI (1986.5.3898.1). Row 3: Los Alamos National Laboratory. © Steve Allen Dreamstime.com. Wikipedia. Los Alamos National Laboratory. Wikipedia.

Full size image

The technical motivation of UO2 as the dominant choice for water-cooled reactors is well-documented and previously discussed. By 1956, it was being used for commercial energy production at the Shippingport Atomic Power Station in Pennsylvania in the United States.

Other types of ceramic fuels have been studied and developed over the decades. For example, UN is the fuel of choice of the BREST-300 reactor under construction in Russia, which is of lead-cooled fast-reactor type. UN was also extensively investigated by the United States for space nuclear power applications in the 1960s and 1970s.23 It is an attractive fuel form being considered by several developers for a range of reactor types, but remains to be qualified and deployed in a commercial setting. UC or UC2–x was the original kernel of TRISO fuels, but has not been utilized since the 1960s, when UO2 and later UO2 + UC2–x emerged as the preferred TRISO fuel kernels. UCx was also investigated as monolithic fuels for gas reactors and for sodium-cooled fast reactor starting in the 1960s. It shares many properties with UN, but exhibits higher swelling that favors the latter in many reactor types. Uranium silicides are being explored in research reactors, and have also been considered for power reactors. For example, it was considered as a replacement for UO2 in light water reactors after the Fukushima accident in Japan, with the benefit of higher thermal conductivity, leading to lower fuel temperatures and less stored energy, and higher fissile density providing economic benefits. However, as with any nonoxide, the potential for adverse reaction with the water coolant is a challenge.24

Particle fuel, which today primarily means TRISO, is the primary form used in pebble-bed reactors and goes back even further. Proposed in 1944 at what is now the Oak Ridge National Laboratory (ORNL), these types of fuels were used25 in the AVR reactor in (West) Germany, the Dragon reactor in the United Kingdom, and the Peach Bottom and the Ultra High Temperature Reactor Experiment (UHTREX) in the United States in the 1960s. Today, many advanced gas-cooled reactor designs rely on the pebble-bed TRISO design. The same type of design is also pursued with a molten salt replacing the gas as the coolant. Further, TRISO fuel is pursued in other types of reactors with the fuel embedded in compacts that are not of particle nature in order to take advantage of superior ability to retain fission products.

Sidebar: The LANL History of Nuclear Reactors and Fuel

Each of us have spent most of our careers at Los Alamos National Laboratory (LANL), and as such, we have some familiarity with the relatively obscure reactor and development efforts at LANL (then known as Los Alamos Scientific Laboratory). Although well-known reactors (e.g., EBR-1) developed by other national laboratories are typically and rightfully cited when describing the history of nuclear energy in the United States, the LANL efforts also played an important role. Specifically, though LANL was established to develop the first atomic bomb, the consequence of research into nuclear reactions led to the development of a number of reactor and fuel designs. LANL has thus played a significant role in the history of nuclear fuel development. As early as 1944, LANL designed and developed the LOPO—Low Power—or Water Boiler reactor, using the nation’s entire stock of enriched uranium as fuel in the form of uranyl sulfate. LOPO was the first reactor to serve as a source of neutrons. This was followed in succession by HYPO and SUPO. HYPO, or High Power, used uranyl nitrate while SUPO (Super Power) operated up to 35 kW.

Clementine [1], going online in 1949, was the world’s first fast neutron reactor and used solid Pu as its fuel, cooled by liquid mercury. Fast burst reactors, Lady Godiva and Godiva II [2], were built and operated beginning in 1951. Lady Godiva utilized a spherical uranium core while a cylindrical design was used for Godiva II. The Ultra High Temperature Reactor Experiment (UHTREX) was a gas-cooled reactor built in 1959. It used UO2 as its fuel and was built to study ways to reduce the cost of nuclear power. And the LAMPRE [3] series of experiments, in the 1950s and 1960s, used liquid Pu as the fuel.

Finally, and perhaps most notably, LANL has been heavily involved in space reactor research. Beginning as early as 1956 and extending into the 1970s, LANL developed reactors for nuclear thermal rockets—some of which were operated in Nevada. In particular, the Rover rocket-propulsion program [4], which operated from 1955 to 1972, used various reactor designs, including Kiwi and NERVA, some of which operated at 1000 MW. These innovative reactor efforts have paved the way for renewed and current interest in not only nuclear reactors for nuclear thermal propulsion [5], but also fission surface power.

[1] H.K. Patenaude, F.J. Freibert, Nucl. Technol. 209, 963 (2023)

[2] T.F. Wimett, R.H. White, W.R. Stratton, D.P. Wood, Nucl. Sci. Eng. 8, 691 (1960)

[3] J.H. Kittel, B.R.T. Frost, J.P. Mustelier, K.Q. Bagley, G.C. Crittenden, J. Van Dievoet, J. Nucl. Mater. 204, 1 (1993)

[4] R.W. Spence, Science 160, 953 (1968)

[5] J. Kelvey, Aerosp. Am. 61, 30 (2023)

The first uranium metal fueled reactor was the X-10 graphite reactor26 at Clinton Engineering Works, today ORNL, which went online in 1943. It was part of the Manhattan Project, and was designed to produce plutonium, but it was also the first reactor to produce useful quantities of electricity. This was followed by EBR-1 in Idaho in 1951, part of what is today the Idaho National Laboratory, which was also fueled by uranium metal. In between these two reactors, Los Alamos National Laboratory (LANL) built a reactor that used pure plutonium metal as its power source. The Clementine reactor at LANL (then known as Project Y) was developed and used during the postwar years of 1945–1952. Clementine used solid plutonium, while LAMPRE I and LAMPRE II, also developed at LANL in the 1960s, used molten plutonium. EBR-II, which operated from 1964 to 1994 at what is now Idaho National Laboratory, used a uranium–zirconium alloy. There are other less common fuel types that were primarily used in research reactors. One example is TRIGA fuel, which is a uranium–zirconium-hydrogen compound.

Molten salt reactors have a long history, dating back to the US Aircraft Reactor Experiment (ARE), part of the US Aircraft Nuclear Propulsion program, in the 1950s. In this case, the salt—NaF–ZrF4–UF4—was also the fuel. Molten salt reactor research hit its peak with the Molten Salt Reactor Experiment (MSRE) at Oak Ridge National Laboratory.27,28 However, by the 1980s, most R&D into molten salts stalled as the overall push to develop nuclear energy also declined. This has changed in the last few years, motivated by private industry and the concept of the small modular reactor (SMR). This includes fluoride-cooled high-temperature reactors (FHRs), where molten salts are used as coolant only with the fuel contained in TRISO particles embedded in pebbles, chloride-based fast-spectrum molten salt reactors with the fuel dissolved in the salt, and fluoride-based reactors with the fuel in the coolant.

As we have seen, many of the fuel forms being pursued today actually go back many decades. Serious investigation of those materials was abandoned or reduced for a number of reasons. Often, simple economics, particularly related to government investment and the lack of public support, is responsible, leaving many unanswered questions that modern efforts can interrogate. However, there are cases where fuel forms were discarded simply because they either proved to have poor properties or they were not economically competitive. The current activity on fuel development might learn from those past experiences, though unfortunately too many results are lost to technical reports that are difficult or impossible to access. The overall fuel R&D enterprise would benefit from these past efforts and it would benefit the community if access to those insights could be better obtained.

The next 50 years

It has barely been 80 years since the first human-controlled sustained nuclear reaction—the entire field of nuclear fuels is not much more than 50 years old. It is thus hard to speculate what the next 50 years might bring. However, it is remarkable to note that after 80 years of nuclear fuel development, we primarily still use one of the original forms used by Fermi—uranium dioxide—in most commercial reactors. Despite the large R&D efforts devoted to the development of improved nuclear fuels, we keep returning to that tried-and-true material. This is to no small degree driven by the fact that the reactor fleet is still dominated by light water reactors, which places a premium on the properties of UO2. We do expect the development and deployment of non-water-cooled reactors to drive adoption of new fuel forms. Also, the fact that the chemistry of the fuel used in light water reactors is still (approximately) UO2 does not mean that the microstructures and thus properties are the same. For example, the density of today’s fuel pellets is higher than in the early days due to improved manufacturing techniques. Further, dopants are added to influence microstructure and mechanical properties. For example, Cr2O3 can be added to decrease porosity and increase the grain size,29,30 which is believed to influence fission gas release31,32 and mechanical properties (pellet-clad-interaction, PCI) favorably.30 Other dopants such as SiO2 are believed to form a glassy phase at grain boundaries, which improves the plastic response of the fuel.33 These examples highlight that having the same fuel chemistry does not by definition equate to the fuel being the same or there being no development. However, the continued dominance of UO2 is still noteworthy.

In the end, it is economics, rather than materials science, that currently drives the selection of nuclear fuels for commercial use and we see no reason for that to change in the future. Only fuel forms that can be economically fabricated and utilized will see widespread adoption. These are likely going to fall into the already defined categories reviewed in this paper. However, just as UO2 has seen refinement through modification in chemistry and microstructure to improve properties over the years, we expect the advanced fuels adopted for advanced reactor application to be improved by similar approaches as reactor technology itself matures after deploying demonstration and commercial reactor units. In order to have widespread use, fuels that are costly to synthesize will need to demonstrate other benefits that compensate for the higher cost.

A large part of the cost of fuel development comes from the qualification process.3 For deployment in a commercial setting, a new nuclear fuel—or even a relatively minor tweak of an existing fuel—must be qualified, a process involving the manufacturing process, fundamental properties, irradiation, post-irradiation examination, and safety testing. The in-reactor irradiations are both costly to perform and challenging to characterize. Further, the relatively limited time any new fuel form sees in-reactor means that there are always large uncertainties on the performance of the material, leading to relatively conservative use—regulators are hesitant to allow burnup past known bounds. That said, there is always more energy to be extracted. Indeed, once uranium dioxide fuels are removed from service, some 94% of the fissile material still remains. That is, we are essentially tossing aside the vast majority of the energy stored in the fuel precisely because we are unsure as to how the material will perform when we push it further.

Thus, there is a big economic driver for pushing fuels to further burnup. This can only happen if the safety margins are better established for these scenarios. However, obtaining data for higher burnups is particularly challenging because of the need for long reactor campaigns to generate those data in the first place.

In our view, one of the biggest developments to come in the next 50 years is the continued integration of advanced modeling and simulation with experimental data to help establish these bounds and to provide the confidence needed to push the burnup of nuclear fuels to higher levels.34 Physics-based models of fuel performance versus burnup will allow for a predictive extrapolation of properties such that we can place confidence in the fuel performance at higher burnup. There has been a rather significant investment into fundamental physics-based modeling of nuclear fuels—in the United States, represented by the Nuclear Energy Advanced Modeling and Simulation (NEAMS) program.22,31,35,36,37,38,39,40 These types of efforts have elucidated the fundamental physical behavior leading, for example, to fission gas evolution and release as a function of burnup. By developing such physics-based models, we can replace empirical models based solely on experimental data and thus enhance the predictive capabilities of fuel performance models. This provides an opportunity to qualify fuels to greater burnup and enhance both the safety and economics of these materials.

Not only do we need better predictions, but we also need to quantify the uncertainty associated with those predictions. However, these models are often based on atomistic calculations such as density functional theory, which do not have quantifiable error (which is not to say they are error-free). At the same time, as information is coarse-grained across scales, there is an inherent loss in fidelity. That is, even if all the physical properties calculated at the atomic scale were perfect, there would be a loss of information as those properties are passed up in scale. This means it is challenging to propagate errors from the fundamental scale to these performance models. Accelerated fuel qualification (AFQ) concepts as illustrated in Figure 3 provide new opportunities to add uncertainties and provide confidence bounds on predictions.2,22 They can also point to new experiments that can help improve confidence on predictions through an integrated modeling-experimental loop. This increases the value of individual experiments while reducing the total number of costly experimental efforts, enhancing the overall qualification process as a result.

Figure 3
figure 3

By integrating modeling and simulation with experimental validation, accelerated fuel qualification (AFQ) offers a path for decreasing uncertainty and increasing safety margins for extended burnup of existing fuels and the licensing of new fuels, increasing the economic viability of fuel development and use, and reducing the time to license new fuels. Figure and concept adapted from Reference 41.

Full size image

Thus, we expect that the biggest changes to come in nuclear fuel R&D efforts will be less in new fuel forms but in our ability to describe and predict the properties of existing fuels. Perhaps this will lead to new fuel types as well, or at least in optimization of the chemistry and microstructure of existing fuel forms, but improving our ability to predict and qualify performance to higher burnup will have a tremendous impact on the efficient operation of nuclear reactors. Being able to confidently predict the evolution of fuel to higher burnup will allow for greater utilization of the fissile content of the fuel, leading to enhanced power generation and the associated economic benefit.

Summary

Although the study of nuclear fuels is now some 80 years old, progress is slow, precisely because these are challenging scientific problems. This is particularly true for nuclear materials, as relevant tests can take many years to generate relevant data to evaluate performance. That said, the initial decades of nuclear energy R&D saw the introduction of many different fuel types that are still being studied. Despite the current flurry in reactor development activity, most of the proposed fuel forms have a long history, first conceived within the first few decades of Fermi’s initial experiment. The predominance of UO2 as the fuel in most reactors is likely to continue for the foreseeable future, with new advanced reactors gradually bringing new fuel types into the market. There are significant opportunities to better utilize and optimize these fuels—to extract more energy from them—increasing both their economic impact and their overall contribution to energy production.