[Image above] Unit 1 of the Perry Nuclear Power Plant in Ohio. Glass and ceramic scientists are heavily involved in the development of stable and manageable approaches to storing nuclear waste. Credit: Nuclear Regulatory Commission, Flickr (CC BY-NC-ND 2.0)

 

As early as 1949, the threat of radioactive nuclear waste was known to the U.S. government. It was not until decades later, though, with the growth of nuclear energy and weapons programs in the mid-to-late-20th century that the public became aware—too late in some cases—of the dangers.

As communities today continue to uncover the full scope and ramifications of poor nuclear waste disposal practices, demands for the safe, stable, and long-term storage of these extremely hazardous substances grow ever louder.

Deep geological repositories have long been seen as a solution to nuclear waste storage. By the time I was a geology student in the late 1980s, research and development were underway on the Yucca Mountain radioactive waste storage facility. Almost 40 years later, however, that facility has yet to be built, and the nuclear waste it was meant to hold continues to accumulate at nuclear power plants and at dedicated “temporary” storage facilities. Instead, a facility in Olkiluoto, Finland, is expected to become the world’s first deep geological repository for nuclear waste when it opens sometime in the next few years.

Creating repositories is only part of the solution to safe nuclear waste disposal, however. Before transportation and storage, nuclear waste must be processed into a stable and manageable form.

When research on stable nuclear waste forms first began in the mid-20th century, scientists knew that some waste types, such as plutonium, remain radioactive for thousands of years. So, this waste would need to be immobilized within a material that takes a very long time to degrade to help safeguard against future exposure.

With this requirement in mind, the possibility of immobilizing radioactive waste in either glass or ceramic materials was recognized very early on. These materials demonstrate chemical durability, impermeability to liquids, and radiation resistance, all essential properties for a long, stable life in waste storage facilities.

Research and development of ceramic nuclear waste forms initially began at Brookhaven National Laboratory in 1953, while research on vitreous nuclear waste forms took place at Massachusetts Institute of Technology between 1956 and 1957. Research on vitrification and ceramic encapsulation continued in the United States, Europe, and Asia throughout the Cold War era, and today both approaches are considered mature and viable options for the long-term immobilization of high-level nuclear waste.

Fission versus fusion reactions and transmutation

All current commercial nuclear power plants are based on fission reactions. The process of nuclear fission involves splitting heavy elements into fragments, which results in the production of some long-lived radioactive elements, such as plutonium.

On the other hand, the process of nuclear fusion involves the forced combination of two smaller nuclei to form heavier elements. Although some radioactive elements are produced and consumed during this process, none of them are long-lived, thus making fusion a safer process than fission, according to the International Atomic Energy Agency.

But forcing atoms together requires more energy than tearing them apart. On Earth, only small-scale fusion reactions have been achieved using the hydrogen isotopes deuterium and tritium to create helium. These reactions required more energy to induce than they produced, making this process infeasible for commoditized energy production.

Besides energy production, fusion reactions could help make fission waste less radioactive. In space, supernova explosions fuse lighter elements into heavier elements to produce the elements between helium and iron on the periodic table. Meanwhile, merging neutron stars produce elements heavier than iron through rapid neutron capture. If these processes could be replicated on Earth (modern-day transmutation), it could help remediate nuclear waste.

Types of glass and ceramic nuclear waste forms

Vitrification, or the transformation of a substance into a glass, is the most common approach to immobilizing high-level radioactive wastes. In this process, nuclear waste is mixed with glass-forming materials, commonly borosilicates and aluminophosphates, and then melted at high temperatures. The molten mixture is then poured into containment vessels, where it cools and solidifies, thereby trapping the radioactive elements in a glass matrix.

Vitrification has a high initial investment cost, high operational cost, and complex technology requiring qualified personnel. So, it is most economically viable at sites that have relatively large volumes of radioactive waste with stable composition.

Similarly to vitrification, nuclear waste can be immobilized in single or polyphase ceramic matrices. These matrices, which often have structures like those of natural minerals, have a wide range of compositions that often fall within the oxide family, including titanates, silicates, phosphates, tungstates, and aluminates. They can be created via cold pressing and sintering, hot pressing, hot isostatic pressing, and spark plasma sintering, among other methods.

Compared to glass matrices, ceramic matrices are sometimes more limited in the type of waste that can be encapsulated due to having a crystalline rather than amorphous structure. Glass-ceramic waste forms can thus offer a compromise that combines the advantages of both glasses and ceramics.

Cements are another option for nuclear waste encapsulation. Cement waste forms are often simpler to create and less expensive than vitreous and ceramic waste forms. However, many questions remain concerning the long-term durability of the cement matrix.

Production of ceramic nuclear waste forms: Challenges and considerations

In general, challenges with ceramic nuclear waste form processing include avoiding radiation exposure for the technicians and preventing further nuclear reactions.

More specifically, the heat required for certain processing methods to create ceramic nuclear waste forms can result in phase transitions within the ceramic matrix, which can weaken the structure and jeopardize its long-term stability. In addition, some radioactive compounds volatilize at high temperatures and can escape from the matrix. Additional off-gas capturing systems are required when processing materials that are prone to volatilization.

On the technical side of processing, some typical ceramic processing methods, such as hot pressing, spark plasma sintering, and solid-state reactions, can be impractical at production scales. Melt processing has recently been proposed as a feasible production-scale formation method for multiphase ceramic nuclear waste forms.

Additionally, some nuclear processes create radioactive salt waste, which cannot be vitrified. Fortunately, salt wastes can instead be blended with zeolite and glass to form a glass-bonded sodalite. This ceramic form traps the radioactive salt inside the zeolite structure and helps to reduce criticality issues as well as making the salt much less soluble.

Finally, although glass and ceramic materials are in general very stable, it is possible that internal and external environmental factors will cause the matrix to degrade before the encapsulated waste is no longer radioactive. So, multiple safety procedures and systems should be considered and implemented in any nuclear waste storage facility.

Machine learning developments in glass and ceramic nuclear waste forms

To assure the public that nuclear power is safe, scientists must be able to demonstrate the long-term stability of vitrified and ceramic nuclear waste forms. Computer modeling and simulations can help with this confirmation by letting researchers understand the microstructure of ceramic phases and how encapsulated radioactive materials will degrade their structures.

Machine learning is increasingly being used as a basis for these models. For example, machine learning models were used to study the leaching behavior of pyrochlore as a potential ceramic nuclear waste form. Other machine learning models were used to study the compatibility of nuclear waste with apatite and hollandite structures or to explore the composition of hollandite-type ceramic nuclear waste forms. Additionally, machine learning models were used to estimate the long-term stability of certain vitreous waste forms.

Ultimately, the use of machine-learning-based methods to develop improved ceramic nuclear waste forms will likely continue to expand, thus supporting  the growth of the nuclear power industry and contributing to lower carbon emissions in the future.

Author

Becky Stewart

CTT Categories

  • Education
  • Energy