Hello. Today, I would like to talk about the recent advancements in nuclear fusion technology and the increasing demand for energy.
First, let me share the good news from March 20th, 2024. A Korean research team succeeded in maintaining a 100 million degree Celsius plasma for 48 seconds in the KSTAR nuclear fusion research device they developed. This breaks the previous world record of 30 seconds. KSTAR already held the previous record as well.
If we can maintain 100 million degrees for 300 seconds, the world will change dramatically as it would lay the foundation for building commercial nuclear fusion power plants. The Korea Institute of Fusion Energy (KFE), established in the Daedeok Research Complex in 1996, has been operating KSTAR, a superconducting nuclear fusion research device using the tokamak method, since 2008.
Other countries' reactors had difficulty with long-term operation due to the high temperatures inside the copper electromagnets. However, KSTAR, utilizing superconducting magnets in its tokamak device, had an advantage in the cooling aspect. KFE's tokamak device became capable of stably maintaining plasma for a long time by generating powerful magnetic fields.
Although Korea started nuclear fusion research relatively late, it was able to quickly catch up by utilizing superconducting magnets. KFE achieved a new world record of 48 seconds in maintaining an ultra-hot plasma over 100 million degrees Celsius, and aims to reach 300 seconds by 2026, which would make 24/7 commercial operation possible.
Private companies are also actively researching nuclear fusion. In May 2023, OpenAI's Sam Altman invested $375 million in Helion Energy, a company trying to generate electricity through nuclear fusion. Altman views providing stable and environmentally friendly electricity as one of AI's tasks.
The demand for electricity in the U.S. is expected to rapidly increase over the next 10 years. The Inflation Reduction Act (IRA) is leading to more energy-intensive facilities and wider adoption of electric vehicles. Particularly, AI has been a major driver of increased electricity demand. By 2026, data center power consumption is expected to exceed 1,000 TWh, with 75% being used for AI.
Traditional renewable energy sources like solar and wind are unstable due to weather conditions, making it difficult to meet such demand. As a result, data centers connected to nuclear power plants are preferred. Amazon's acquired Talen Energy data center is linked to a nearby nuclear plant, and Microsoft and Google are also pursuing small modular reactors (SMRs).
Helion Energy, where Sam Altman invested, aims to commercialize nuclear fusion power within 5 years to meet this electricity demand. Nuclear fusion generates energy by fusing light atomic nuclei like deuterium and tritium to form heavier nuclei, based on Einstein's mass-energy equivalence principle.
When deuterium and tritium collide, they form a helium atom, with the mass deficit released as energy along with neutrons. These neutrons hit the reactor walls, generating heat to power the plant, similar to how the sun operates through nuclear fusion as a massive reactor.
However, forcing nuclei to fuse requires overcoming electrostatic repulsion to bring them close enough for nuclear forces to take over, requiring immense pressure and energy. Pressures as high as the sun's core are needed, which necessitates temperatures over 100 million degrees Celsius.
Heating deuterium and tritium above 100 million degrees produces a plasma state where atomic nuclei and electrons are separated. In this state, the nuclei can fuse and release energy. The challenge was finding materials to contain such ultra-hot plasma.
In 1952, Soviet scientists Igor Tamm and Andrei Sakharov developed a doughnut-shaped magnetic coil chamber to confine plasma, known as the tokamak. An internal current creates the plasma, which follows the magnetic field of giant coils while circulating in the doughnut, triggering fusion reactions.
Superconductors are crucial for generating and sustaining powerful magnetic fields with zero resistance. Although current superconductors require extreme cooling, the development of room-temperature superconductors would greatly accelerate nuclear fusion commercialization.
If the plasma meets certain conditions, it can sustain nuclear fusion without external heating, becoming an artificial sun. In 1951, Lyman Spitzer of Princeton proposed the stellarator theory, a twisted, pretzel-shaped plasma confinement device.
Stellarators can contain plasma more stably than tokamaks for longer periods but have complex structures that are more expensive to manufacture. While stellarators were initially favored, advancements in precision control technologies like AI have addressed tokamak drawbacks, making them more cost-effective. Tokamaks are currently considered the most commercially viable approach.
To commercialize tokamak-based nuclear fusion, the U.S., China, Russia, the EU, Korea, Japan, and India are collaborating on the International Thermonuclear Experimental Reactor (ITER) project. Korea is contributing 9% of the construction costs, with major companies securing over $7 billion in orders. ITER's 80m² super-large device aims to demonstrate that nuclear fusion can be a viable energy source.
If ITER succeeds, it could achieve breakeven where the energy output exceeds the input. The first plasma ignition is scheduled for this year, with key experiments expected around 2027. Running ITER annually requires 12kg of tritium, potentially up to 18kg cumulatively.
There were suggestions to use tritium from contaminated water at Japan's Fukushima nuclear plant for ITER, as tritium has similar properties to water, making separation difficult. However, the total amount of only 2.2g was deemed insignificant. Currently, tritium is commercially supplied from Canadian nuclear plants.
Aside from Canada, Korea's Wolsong nuclear plant also produces tritium, with 5.7kg in storage. Tritium is currently used in small amounts for items like glow-in-the-dark products – 200-300 million becquerels in glow watches and 9 trillion becquerels in emergency exit signs.
In March 2024, Korea's Nuclear Safety Commission approved expanding tritium transportation and storage containers. With tritium prices 400 times higher than gold, Korea is considering exporting its stockpile. This implies the tritium in Japan's contaminated water also has value.
There are two main approaches to nuclear fusion: using tritium or deuterium+helium-3. While deuterium is easily obtained from seawater, tritium requires nuclear conversion of lithium. Although tritium fusion has a shorter half-life, making it easier to manage than uranium, the helium-3 from the lunar surface is more economically viable with deuterium.
However, the tritium fusion approach cannot be completely discarded. While helium-3 from the moon is a long-term solution, tritium-based fusion power will play an important role in the short term to meet immediate electricity demands.
The U.S. is making an all-out effort to secure helium-3 from a lunar base. Helium-3, brought to the moon's surface by the solar wind over meters of depth, has immense potential. Just 1 ton of helium-3 can generate as much energy as 14 million tons of oil, with 25 tons enough to power the entire U.S. for a year. The moon's surface contains enough helium-3 to supply humanity's energy needs for 10,000 years.
This is why the Starship launched on March 15, 2024 aims not just to land on the moon, but to establish a permanent presence for mining helium-3 and rare earth elements. The massive 122m Starship, developed for crewed Mars exploration, combines the 50m Starship and 70m Super Heavy rocket in a two-stage vehicle.
The Super Heavy booster is twice as powerful as NASA's SLS, making it the most potent launch vehicle in human history. Due to its capabilities, the Starship was also selected as NASA's Artemis lunar lander, unlike smaller previous landers.
Recent progress in nuclear fusion technology seems to be outpacing expectations. While tritium-based fusion has limitations, the deuterium approach utilizing the moon's helium-3 is expected to lead commercialization efforts. However, it may be too soon to meet the immense power demands of AI and other applications.
As a result, small modular reactors (SMRs) are emerging as an interim solution until fusion is commercialized. Bill Gates' TerraPower has already obtained U.S. regulatory approval and begun construction, targeting the first SMR operation by 2030. In addition to $1 billion in private funding, TerraPower will receive $2 billion in government support.
The $1 billion in private funding includes $250 million from SK and SK Innovation in Korea, while Hyundai Motor Group secured a 10% stake with a $30 million investment. In March 2024, NEMO, a maritime nuclear power consortium involving SMR companies and nuclear plant operators, was also launched.
NEMO aims to establish global standards and regulations for commercializing maritime nuclear power. Participants include TerraPower, Westinghouse, Seaborg from 7 countries and 11 companies. Notably, Seaborg from Denmark possesses molten salt reactor technology using sodium, considered suitable for maritime applications.
Korean companies are also actively involved in this arena. As electricity demand grows exponentially, the ability to stably supply sufficient power is becoming a critical factor for national competitiveness. New opportunities are expected to emerge across the entire process of generating, transmitting, and storing electricity.
While researching nuclear fusion, attention naturally shifts towards SMRs, as fusion may not meet short-term power demands even if realized. With various approaches being explored and competing, it remains to be seen which technology will take the lead.
You may also be interested in:
Starship 3rd Launch Update (feat. Elon Musk, Artemis)
Unraveling History: A Different Lens on the Russia-Ukraine War
