Korea's Nuclear Expansion Rests on Inflated Demand Projections, Activists Say
A Korean civic coalition is challenging the government's latest electricity demand forecast, which surged by 28GW in months after industrial mega-projects were announced. The group argues the figures are speculative and being used to justify new nuclear plants that can't arrive in time anyway.
The Numbers Don’t Add Up
South Korea’s government recently unveiled its 12th Basic Electricity Plan, and the headline figure tells a story that should make anyone skeptical. The projected additional demand from semiconductors and data centers has exploded to more than 36GW — up from roughly 8GW in April, an increase of about 28GW in just a few months.
This isn’t a gradual revision based on new data. It’s a sudden leap that coincides almost exactly with the June announcement of three industrial mega-projects by President Yoon’s administration. The government confirmed the methodology plainly: when companies filed investment plans, the ministry treated them as settled demand and reflected “the full amount” in its projections.
That is a policy choice, not a scientific conclusion. And it is the foundation upon which Seoul now wants to build a wave of new nuclear reactors.
How the Logic Works — and Why It Fails
The sequence is transparent. The government and industry agree on massive investment projects. Those projects generate enormous electricity demand on paper. The government then calculates how many nuclear plants are needed to serve that demand. The result justifies both new reactor construction and extensions for aging units that may otherwise face mandatory shutdowns.
The civic group driving the challenge, the Nationwide Action for Stopping New Nuclear Power Plants and a Nuclear-Free Society, puts it starkly: this approach calculates how many reactors are necessary rather than asking whether the projected demand is real. The logic runs in the wrong direction — demand is assumed first, supply is calculated second, and nuclear becomes the default answer regardless of whether the demand itself is justified.
The Speculative Core of the Forecast
The projections rest on facilities that do not yet exist. The 24.3 to 24.6GW attributed to advanced industries and the 11.9GW attributed to data centers represent planned capacity, not operational load. Construction has not begun on many of these sites. Investment decisions can shift with market conditions, technology changes, or corporate strategy.
Treating speculative demand as if it were locked in creates a self-fulfilling prophecy. Inflated projections create an artificial shortage, and the shortage is then used to justify building nuclear plants that were always meant to fill a gap of the government’s own making.
The activist group’s statement, released August 20, 2026, argues this is not a minor methodological dispute. It is the central fault line in Korea’s energy policy. If the demand is overstated, then the entire case for nuclear expansion collapses with it.
The Timing Problem
Even if the demand projections were accurate, there is a glaring mismatch between when the electricity would be needed and when new nuclear plants could deliver it. The government plans for semiconductor clusters and AI data centers to begin full operation in the late 2020s. A new large nuclear plant — from site selection and permitting through construction and commissioning — requires at least a decade. The clocks do not align.
This means the industrial power demand cannot realistically be met by the new reactors the government is now proposing. The plants being planned will come online well after the initial surge in demand has either been met by other sources or rendered unnecessary by shifting technology and efficiency gains.
Yet the projections built for 2040 are being used to greenlight construction decisions today. A forecast fabricated over three months is becoming the justification for infrastructure that will operate for decades.
The Generational Gamble
Beyond the timing issue lies a deeper risk. The government is not only talking about conventional large reactors. It is now also pushing to introduce fourth-generation reactors — non-water-cooled SMRs and MMRs — for deployment after 2035. These technologies have not been commercially validated. Their economics, safety profiles, and waste management challenges remain unresolved.
The demand projection has become so expansive that it is being used to pull unproven reactor designs into the supply mix as well. That is a policy built on a stack of uncertainties: inflated demand, unproven technology, and a timeline that does not match reality.
The Geographic Injustice
There is also a distributional question that rarely appears in energy policy debates. Nuclear plants and high-voltage transmission lines do not benefit the regions where they are built. The semiconductor complexes and data centers drawing power from these plants are concentrated in industrial zones and the capital region. The risk — of accidents, of nuclear waste, of infrastructure intruding on communities — falls on the host regions.
The civic group describes this as unjust energy transition. Companies pursue mega-projects in one location, nuclear plants and transmission corridors are sited elsewhere, and the costs accumulate in communities that had no say in the decision. The benefits flow to Seoul and the tech corridors; the burdens stay local.
What Should Happen Next
The immediate question is whether the 12th Basic Electricity Plan’s demand assumptions will survive scrutiny. The civic group is calling for a re-examination of the projections — not as a technical formality but as a genuine audit of what is real demand, what is speculative, and what is double-counted.
If Korea follows the path the government currently outlines, it will lock in new nuclear capacity based on numbers that may not materialize. The commitment will outlast the political cycle and the industrial boom that justified it. Other countries watching this debate — particularly those considering nuclear expansion to serve AI and semiconductor demand — should note the methodological gap.
The alternative is to start from the opposite premise: set climate targets, assess how much electricity can be saved through efficiency and demand management, evaluate where industrial growth actually requires new capacity, and build supply to meet that verified need. That approach is harder. It requires discipline. But it is less likely to produce a fleet of reactors built for a shortage that was invented.
The 36GW figure is the number that matters. It is large enough to reshape Korea’s energy system. It is also fragile enough to be wrong. Whether it is an accurate reflection of future demand or a political construct dressed in technical language will determine what kind of energy system Korea inherits — and who pays for it.