Faster testing – better nuclear
Good engineering is often an evolutionary process. You try an idea, model it, build a prototype and see, if it works. More often than not, something will break down or not really work as anticipated.
Then, it is back to the beginning, but with more information. The faster a company is with each of these iterations, the better they tend to do in the market.
The nuclear sector in the Western world works in a regulatory framework, that has in a lot of ways become the antithesis of this approach.
What can we do about it?
Nuclear energy development has been brought to a near standstill sometime in the 70s. One of the major obstacles has always been the rather long time it takes to qualify materials and fuels to the regulators’ satisfaction.
Recently, there have been multiple initiatives to accelerate the testing of nuclear materials.
I’d like to highlight what I believe to be the most interesting routes.
Going digital
Simulation Models, especially the ones unifying different physical domains, like neutronics, hydraulics and thermodynamics.
While there have been multiple specialized physics kernels for decades, a growing movement towards making them available either via open source or at least for paying service fees is under way.
TerraPower’s team joins developers of similar open-source software, such as INL’s MOOSE (Multiphysics Object-Oriented Simulation Environment), MIT’s OpenMOC and OpenMC and others, after releasing parts of the advanced reactor modeling interface, to work collaboratively with DOE and the nuclear industry in their efforts to develop and demonstrate advanced reactor technologies. The open-source portion of TerraPower’s software offers a framework and ecosystem that will benefit from external contributors who are also in need of better automation tools and expressive reactor frameworks.
With an open-source portion of TerraPower’s software, TerraPower’s advanced reactor modeling interface provides an integration platform that can improve the usability and coupling of physics kernels and reactor analysis methodologies, whether they are proprietary or publicly available.
It is hoped that universities and maybe even startups can contribute to these projects or provide data for the libraries. An international movement to extend and validate the data bases and, even more critically, to provide a common framework for entrepreneurs and regulators. TerraPower’s advanced reactor modeling interface software interacts with the powerful proprietary physics kernels of the U.S. Department of Energy labs and other vendors, such as TerraPower’s own kernels. These have still to be purchased.
Together with these proprietary kernels, this platform allows engineering-level design and analysis of myriad different processes for generic reactor designs. It also reduces the amount of data management that needs to be done in the beginning stages for most companies. Design parameters must be provided by the user. This allows national laboratories, regulators, universities and reactor developers to evaluate concepts at much earlier stages of development.
Experience from other sectors with open source software indicates that this project could help to shave off years of effort for new companies to model their designs and might enable them to skip building the software necessary to do so even entirely.
Material:
Building materials are of special importance in the nuclear sector. The conditions inside a nuclear reactor are among the most extreme environments on earth. The levels of neutron radiation, the heat and depending on the type of reactor, the presence of potentially corrosive substances, like lead or fluoride salts, makes testing these materials a challenge. Especially the reactor vessel is usually designed to last for decades, because a replacement is typically technically and economically challenging.
To simulate the extreme conditions inside a nuclear reactor, Argonne has developed a tool, which is partially microscope and partially particle accelerator.
In it, a whole year of reactor radiation can be simulated in just one day, which should accelerate testing quite a bit.
Elysium Industries and Moltex Energy have their own fuel production processes.
A major factor for the lengthy testing regimes of solid fuels is the structural, radiological and chemical integrity of the cladding and fuel pellets.
With a move towards liquid fuels, there are no claddings and no solid structure to be maintained. The chemical properties of the fuel can often be simulated with stable or only mildly radioactive isotopes.
Because fission gases are constantly removed there is no pressure build up, which has to be contained. There is nothing comparable to the disassociation of water catalyzed by hot zirconium cladding in MSRs. On the other hand, the chemistry becomes more challenging, because so many chemical species get created during the fission and decay process.
Producing and qualification of liquid fuels is expected to be significantly cheaper and faster than for solid fuel.
Integrated testing:
Terrapowera and GEH are developing a the new Verstaile Test Reactor, which will enable the qualification of materials and fuels for fast reactors. It is a sodium-cooled, solid-fueld reactor. It stands to reason that this reactor will share a lot of commonality with the “Natrium” system, a reactor system developed by these very same companies, which recently got a $80M investment from the DoE.
TerraPower is also partnering with Southern Company to build an integrated testing facility for the molten chloride fast reactor (MCFR). The DoE has already invested more than $28M for the project to identify and test materials used in the reactor.
The MCFR offers significant safety, performance and economic benefits. It is walk-away-safe and does not need any electric pumps to prevent fuel damage. There are no fuel assemblies to fabricate, replace or store. It is refueled continuously during operation, can use a broad spectrum of fuels and automatically load follows.
Summary
The tendencies at work here are obvious: test often, small and especially fast and with bounded risk. I am not aware of any branch of technology, where this has not helped to improve the technology. If this trend persists, fast nuclear innovation cycles lie before us.
This might give us a shot a climate change!