Reproduced from medium.com
While engineering for Mars and space exploration could spur significant advances in science and technology in a techno-utopian world, it’s human nature to prioritize basic needs like security before exploration and conflict is an eternal part of the human condition. In 2022, the U.S. Department of Defense’s budget was $344 billion and the U.S. Department of Homeland Security’s budget was $52 billion, which together are 16 times greater than NASA’s budget of $24 billion. In addition, almost half of the U.S. Department of Energy’s $48 billion budget was devoted to maintenance of the nuclear weapons stockpile, non-proliferation activities, and military-grade naval nuclear propulsion reactor design and operations.
Many of the latest advances in artificial intelligence, robotics, and even some electric power plant technology have military roots. Computer vision has achieved significant advances not only through self-driving cars and internet consumer image processing, but also through use in unmanned drones to detect and destroy military targets. Natural language processing is not only used to develop customer support chatbots, but also to organize military intelligence sources. Humanoid robots are not only being developed for civilian applications like manufacturing, but also for battlefield operations (as well as more controversial domestic applications like police department operations using surplus military equipment). Before the latest advances in AI and robotics, the military’s heavy hand also guided the direction of nuclear power plant development — and shaped the choice of using uranium fuel vs. alternatives like thorium, with consequences for the health of miners and placing an eternal burden on the U.S. federal budget.
Uranium nuclear fuel cycle
The United States currently has 93 commercial nuclear reactors in operation around the country. Together these reactors provide about 20% of the nation’s electricity and about 50% of the carbon emissions-free supply. Of the 93 nuclear reactors in the U.S., 62 are pressurized water reactors (PWR) and 31 are boiling water reactors (BWR) providing 95,492 MWe of electric capacity. In order to power the U.S. nuclear fleet, every year between 2–38 million metric tons of natural uranium ore is mined, milled into uranium oxide concentrate (also known as yellowcake), converted into uranium hexafluoride and enriched to 3.5–5% of the uranium-235 isotope suitable for nuclear fission, and then fabricated into approximately 2000 metric tons of uranium oxide fuel pellets for use in a nuclear reactor. In comparison, making electricity with coal, which also supplied ~20% of U.S. electricity in 2021 requires mining approximately 454 million metric tons of coal ore, requiring 10 times more resource mining than uranium. This is because when an atomic nucleus decays through fission, it releases radiation proportional to the strong nuclear force which binds protons and neutrons together, which is about 100 times stronger than breaking intermolecular forces through the combustion of coal. Like most types of metal mining, uranium mining has environmental risks due to exposure of radioactive gas and dust or mine tailings in the air and water.
Uranium ore in its most common isotope, uranium-238, does not pose health hazards unless it is inhaled or ingested and has a similar level of radioactivity as granite. But uranium decays into radon gas, which decays into radioactive isotopes of polonium and lead which do pose an inhalation risk at the concentrations found in mines. These risks can be mitigated with personal protective equipment, mechanical ventilation systems and limiting exposure times. These risks and corrective actions were not initially disclosed to uranium miners during the early stages of the U.S. nuclear power industry under federal government control because scientists’ warnings were minimized or uncertainties in the science paralyzed taking any corrective action or information-sharing with the mining community. Through the advocacy of miners’ families, union and tribe leaders, including from the Navajo people, the Radiation Exposure Compensation Act (RECA) was passed in 1990 to address the harms to uranium miners.
The most complex part of the nuclear fuel production process is fuel enrichment. Naturally occurring uranium contains only 0.7% of the fissile uranium-235 isotope so mined uranium must be processed to separate isotopes and create an enriched fuel that contains 3.5–5% of uranium-235. Uranium enrichment is primarily conducted with centrifuges, consisting of vertically spinning tubes that separate uranium hexafluoride gas by weight of isotopes. The uranium enrichment process contains a number of workplace hazards, as it involves handling the radioactive and chemically toxic uranium hexafloride, which can react with moisture to create the highly toxic hydrofluoric acid, before it is chemically converted into a uranium dioxide metal for a fuel rod. The international nuclear industry has workplace standards to manage radioactivity, chemical spills, mechanical failure or other accident risks.
Thorium nuclear fuel cycle
Early in the development of commercial nuclear power plants in the U.S., physicists and nuclear engineers knew of an alternative nuclear fuel cycle that would require much less mining, fuel, and water and produce cheaper electricity than uranium. This fuel cycle uses element 90 in the periodic table, thorium, which is 3–4 times more abundant on Earth compared to element 92, uranium. According to Physics Nobel Prize winner Carlo Rubbia, who also holds a patent in a proton accelerator-based thorium reactor, every unit of thorium is equivalent to 200 units of uranium in order to produce electricity. This is because the thorium nuclear fuel cycle relies on the most commonly occurring stable isotope found on Earth, thorium-232, would have required limited fuel enrichment.
The thorium fuel cycle is shown below. When thorium-232 absorbs a neutron, it transmutes into thorium-233 and decays, releasing a beta particle to protactinium-233, which further delays into uranium-233, which can be used to continue the cycle. The thorium fuel cycle has the advantage of being a ‘breeder’ reactor, which means each cycle produces multiple neutrons to continue the fuel cycle. However each thorium fuel cycle takes 27 days and 22 minutes compared to the standard uranium-238 fuel cycle, which has a ‘cycle time’ (it’s actually a half-life decay period, but it is somewhat analogous to a production cycle time) of under 3 days. The thorium fuel cycle can be designed to be closed loop, producing limited to no long-term nuclear waste products with 1000+ year half-lives.
Thorium is obtained from processing the monazite mineral, which is also a good source of lanthanide rare earth materials. Monazite handling does have a number of hazards and protection requirements, primarily due to any inhalation or ingestion of ore dust that might occur through workplace exposure, similar to uranium mining. Using Rubbia’s material requirements estimate, a thorium-based nuclear fuel cycle could potentially provide 20% of U.S. electricity with approximately 200,000 metric tons of thorium ore. This is 0.04% of the mining requirements of coal-powered electricity or 0.5% of the mining requirements of the uranium-powered nuclear fuel cycle. So why did the U.S. not pursue this potentially cheaper, more environmentally-friendly, and more energy-efficient nuclear pathway?
The choice of the uranium fuel cycle in the U.S. nuclear power plant industry is a classic example of the path dependence of technology development. In the post-World War II era, much of the nuclear research and engineering work required to design and develop complex uranium fuel enrichment technology for the Manhattan Project nuclear weapons program was then transferred to the ‘Atoms for Peace’ program by President Eisenhower. This includes the only currently operating gas centrifuge uranium fuel enrichment facility in the United States known as URENCO USA in Eunice, NM. The first commercial nuclear power plant built in the U.S. was a joint effort led by Admiral Hyman Rickover, drawing on his experience leading the development of naval nuclear propulsion systems, and the Westinghouse Electric Corporation to design and build the world’s first PWR nuclear power plant at the Shippingport Atomic Power Plant in Beaver County, Pennsylvania. Private U.S. companies then continued down the uranium fuel cycle path, with regulatory oversight, and went on to build the nuclear power fleet that exists today. There was once a thorium molten salt reactor program out of Oak Ridge National Laboratory, led by Alvin Weinberg, but it was dismantled in the 1970s due to perceived nuclear proliferation risks.
Today, some of the latest advances in the industry have been in the development of small modular nuclear reactors, in part because these are the only types of power plants that can economically be built by private companies without large federal loans or other subsidies. Traditional 1000s MW-scale nuclear power plants like the Alvin W. Vogtle plant, with 4 planned reactors and 2 currently under construction, have a cost of approximately $17 billion in today’s dollars including a $3.7 billion loan guarantee from the U.S. Department of Energy. Small modular nuclear reactors are easier to finance in a market-based private electric power industry, as is found in the U.S. For example the 77 MWe NuScale Power Module at a cost of $4,200 per kilowatt would have a total cost between $300–400 million. The vast majority of small modular reactor projects are pursuing a uranium fuel cycle. However, a thorium fuel cycle is also suitable for small modular reactors. This modular design approach is a game-changer from a business financing perspective if the engineering and operational risks can be identified, managed, and carefully regulated.
There is a small but vocal group of thorium enthusiasts in the U.S. including the ThorCon International company, which builds a hybrid thorium/uranium liquid fuel molten salt reactor for domestic use or export. The company recently signed an agreement with Bureau Veritas to build a ‘floating’ molten salt reactor in Indonesia. Flibe Energy is also developing a thermal spectrum lithium-fluoride molten salt thorium reactor (LFTR) and is currently in development stage. The LFTR reactor has the advantage of being able to use a higher temperature Brayton nitrogen generator rather than traditional steam generators, raising thermal efficiency from 35% to 50% and is thought to have lower proliferation risks. Internationally, India has a thorium-based nuclear program with a goal of providing 30% of the country’s electricity by 2050. Canada’s CANDU reactors are capable of using thorium fuel and has been in discussions to build thorium-based reactors in Chile and Indonesia. In addition, Norway’s ThorEnergy has plans for a four-year experiment with thorium fuel in an existing nuclear reactor. And so, thorium-based reactors have a new opportunity to emerge.