South Korea's flagship effort to build one of the world's largest AI and semiconductor manufacturing hubs could be crippled by a lack of electricity unless the country rapidly upgrades its grid and reforms its power markets. New analysis from Wood Mackenzie finds that the country's KRW 4,700 trillion (US$3.1 trillion) national mega-project, anchored by a massive semiconductor and AI data centre cluster in the Honam region, faces a growing power supply crisis that could derail the entire initiative.

The numbers are stark. The Honam cluster alone is expected to push regional peak electricity demand from an estimated 11.8 GW today to 19.1 GW once fully developed, according to Wood Mackenzie. But under current plans, effective regional generating capacity will reach only 16.8 GW by 2030, leaving a shortfall of approximately 2.3 GW. Closing that gap through solar generation alone would require roughly 19.5 GW of new installations within the next five years—a nearly impossible task given that only 679 MW of solar capacity was added across the region during 2025. Meanwhile, about 3 GW of the Hanbit Nuclear Power Plant's 5.9 GW fleet is scheduled to retire by 2035, threatening to widen the gap further. Honam has the country's strongest renewable energy potential, but much of that capacity can't reach consumers because the transmission network is already congested during spring months, preventing new solar projects from securing grid connections.

"South Korea is planning for an entirely new class of industrial load now," said Yunsik Chung, research analyst for Asia Pacific power and renewables at Wood Mackenzie. "The challenge is less about whether enough generation can ultimately be built than whether the grid, market design and investment signals can evolve quickly enough to support AI infrastructure on the timetable policymakers and industry expect." The report concludes that rapidly expanding battery energy storage systems alongside market reforms will be essential, but current market design remains one of the largest barriers—batteries can presently generate revenue primarily through South Korea's central contract market, limiting commercial deployment. Chung added that "this is ultimately a sequencing challenge," warning that if near-term flexibility measures, firm generation, and transmission upgrades don't move in parallel, "the pace of AI and semiconductor investment could become constrained by power infrastructure rather than technology or capital."

The report identifies several solutions rooted in the disconnect between South Korea's explosive AI-driven industrial growth and an electricity system built for a different era. The Honam region's transmission corridor to central South Korea is already congested, meaning renewable projects can't connect even when resource availability is abundant—a bottleneck that requires both physical grid expansion and non-wired alternatives like dynamic line rating to relieve pressure while larger transmission projects are developed. Market reforms are equally urgent: without either substantially expanding procurement volumes or introducing new revenue streams for battery storage, the economics won't support the pace of deployment needed to balance intermittent solar generation. On the supply side, extending the Hanbit Nuclear Power Plant's operating life by 10 years could preserve roughly 2 GW of firm capacity that would otherwise retire, while prioritizing domestically manufactured gas turbines could sidestep global supply chain delays that currently stretch five to six years. Over the longer term, brownfield nuclear development on land reserved for cancelled Hanbit units could cut construction costs by 10–15% and shorten timelines, while reinforcing the transmission link to the Yeongnam region—where surplus nuclear capacity exceeds peak demand by 15.9 GW—could funnel excess power to Honam's growing industrial load.

The report warns that if power infrastructure lags, South Korea's ambitious AI and semiconductor push could stall not for lack of capital or technology, but because there isn't enough electricity to turn the machines on. Two structural solutions offer long-term relief: brownfield nuclear development and transmission corridor upgrades that move surplus generation from other regions. But in the near term, every element—grid flexibility, firm generation, and market reform—needs to advance together. As new industrial demand increasingly consumes renewable generation within Honam itself, the scale required for the proposed 8 GW West Coast HVDC transmission project may shrink, with further clarity expected when the government publishes its 12th Basic Plan for Electricity Supply and Demand.