Will the Nuclear Renaissance Driven by SMR Become the Power Crisis Solver in the AI Era? [AK Radio]

One of the most contentious issues in the recent U.S. presidential election was energy policy. Donald Trump, the 47th president-elect, emphasized fossil fuel policies, but his stance on nuclear power did not differ significantly from that of the current administration under Joe Biden. Traditionally, the Republican Party has supported the fossil fuel industry, but bipartisan consensus appears to be forming around nuclear power.

The Biden administration recently announced an ambitious plan to expand nuclear power generation capacity to 300 gigawatts by 2050, tripling the current capacity. The current nuclear power capacity in the U.S. stands at approximately 100.6 gigawatts. Through the "ADVANCE Act" (Accelerating Deployment of Versatile Advanced Nuclear Technologies), the administration is demonstrating strong commitment to fostering the nuclear industry by streamlining complex licensing procedures and easing regulations. This is notable as it sets a long-term energy policy direction despite being late in the administration’s term.

In the U.S., efforts are underway to restart nuclear plants that have been offline since the Three Mile Island accident in 1979. The Three Mile Island plant had been dormant since 2019 but was recently approved for reactivation. This reflects a growing recognition that nuclear power is indispensable for resolving the country’s power shortages.

Amid these trends, Small Modular Reactors (SMRs) are emerging as next-generation nuclear technology. SMRs miniaturize components of traditional nuclear plants, placing them into a single container, which is prefabricated in factories and then assembled on-site.

The greatest feature of SMRs is safety. Unlike large conventional nuclear plants, which risk accidents if power is lost and coolant circulation stops, SMRs adopt a "passive design" that allows natural circulation cooling even when power is cut off. This is regarded as a groundbreaking technology capable of preventing major accidents like the Fukushima nuclear disaster.

Big tech companies are also showing strong interest in SMRs. Bill Gates, founder of Microsoft, established TerraPower in 2006 to lead SMR development. Recently, Amazon has invested in three SMR-related companies and announced plans to purchase power, while Google has also expressed intent to buy SMR-generated electricity. Peter Thiel, a prominent Silicon Valley investor, has invested in companies developing SMR nuclear fuel, indicating ongoing participation from technology firms.

Notably, Chris Wright, nominated as Secretary of Energy under the Trump administration, is a board member of Oklo, an SMR company. Oklo is also backed by Sam Altman, co-founder of OpenAI. This demonstrates that SMRs are recognized as a future energy solution regardless of political affiliation.

Korean companies are actively entering the SMR market as well. For example, Doosan Enerbility and DL E&C are participating in the X-Energy project, which Amazon has invested in. Korean firms are particularly noted as partners for U.S. companies due to their extensive experience in nuclear plant construction and operation. While U.S. SMR startups possess core technologies, they often lack practical construction and operational experience, an area where Korean companies excel.

This interest is also related to the surge in power demand in the AI era. AI data centers consume enormous amounts of electricity?not only for computing operations but also for cooling. Goldman Sachs projects that power demand from AI data centers will increase by 160% by 2030 compared to current levels.

SMRs have the advantage of being installable near data centers. Conventional large nuclear plants must be built on coastlines due to safety and cooling water issues, but SMRs’ relatively higher safety allows installation near power consumption sites. This significantly reduces transmission costs.

However, there are concerns that commercialization of SMRs is still a long way off. TerraPower has not succeeded in commercializing SMRs for over 15 years, and NewScale, a leading U.S. SMR company, recently canceled a project due to rising costs. Initially, SMRs were expected to reduce costs through miniaturization, but instead, economies of scale have diminished, causing an increase in power generation costs.

Moreover, no matter how much safety improves, it remains difficult to fully alleviate residents’ concerns about nuclear plants. Even if SMRs are safer than conventional reactors, psychological resistance to having reactors installed near residential areas persists. This remains a significant social challenge for SMR expansion.

Experts point out that while SMRs are promising technology, they still require validation. Issues such as efficiency loss and cost increases due to miniaturization need to be addressed. Especially with the surge in data center power demand in the AI era, there are concerns that SMRs, which may only be commercialized after 2030, will not be able to meet immediate power needs.

Ultimately, SMRs represent a promising alternative to solve future energy problems, but many practical challenges remain. Experts advise that rather than overly optimistic expectations for SMRs, a comprehensive energy policy combined with feasible short-term measures is necessary.

Governments and companies should continue SMR development while carefully addressing practical challenges one by one. At the same time, short-term solutions to meet the rapidly increasing power demand must be prepared. In the end, future energy issues cannot be solved by any single technology or policy; balanced development of diverse energy sources and technologies will be crucial.