On May 17, 2025, Taiwan shut down the last operating reactor at the Maanshan (Nuclear III) plant — four days after an opposition-led amendment to the Nuclear Reactor Facilities Regulation Act opened a path for 20-year extensions and spurred a failed restart bid.
The clash raises a hard question. Is reviving nuclear power a pragmatic answer to energy security and industrial needs, or a risky bet in a seismic hotspot with aging infrastructure and no final waste solution?
Advocates frame nuclear as essential to both economic vitality and national security. Taiwan’s chip-driven economy — and the AI surge behind it — needs stable, round-the-clock baseload that intermittent renewables cannot yet guarantee on their own.
As Nvidia CEO Jensen Huang said in Taipei, “Taiwan should absolutely invest in nuclear,” casting it as critical to industrial growth.
Security arguments go further. Taiwan imports more than 96% of its energy, a vulnerability in any supply disruption or blockade. Nuclear fuel can be stored on site for 18 to 36 months, offering a strategic buffer. By contrast, liquefied natural gas reserves cover roughly 11 to 14 days at current levels.
From this view, nuclear is more than a power source — it is part of national defense, keeping critical infrastructure running in a crisis.
Opponents counter that large, centralized plants are “soft targets.” An attacker would not need to breach a reactor core; damaging grid connections or cooling-water intakes could trigger cascading risk. The 2022 crisis at Ukraine’s Zaporizhzhia plant is a sobering precedent showing how a nuclear facility can be turned into a strategic pawn and a regional danger during conflict.
They argue true resilience lies in a distributed grid — thousands of rooftop solar arrays, offshore wind, and battery storage — systems that are harder to disable with a single strike and quicker to restore after damage.
Cost is the other battleground. Supporters say the economics are compelling because capital costs have long been amortized. Cited 2025 figures put nuclear generation near NT$1.87 (US$0.06) per kWh, far below externally purchased renewables at about NT$5.60 per kWh, easing pressure on regulated tariffs and on Taipower, whose deficits exceed NT$400 billion due to high fuel costs and capped retail prices.
Critics call “cheap nuclear” a half-truth that ignores the back end. Decommissioning and waste management are immense, and Taiwan’s Nuclear Backend Fund faces a projected shortfall of hundreds of billions of Taiwan dollars. Restarting old reactors also is not inexpensive. Life-extension demands major safety upgrades, replacements, and inspections.
They point to California’s Diablo Canyon — a similar-vintage plant — where estimates to extend operations climbed from US$5.2 billion to more than US$11.8 billion (about NT$380 billion). Applying similar realities in Taiwan suggests life-extension is a massive new investment in aging technology, not a cut-rate fix.
Safety arguments cut both ways. Proponents cite post-Fukushima retrofits, decades of operations without a major accident, and international peer reviews as evidence of robust standards. They add that engineering upgrades have raised assessed earthquake resilience far above original design levels.
Opponents focus on geography and age. Taiwan’s reactors sit near active faults; Maanshan is about 700 meters from the Hengchun fault. A widely cited 1.38 g figure reflects a reassessment of potential hazard, they note, not proven plant capability, and even the best retrofits cannot fully mitigate ground rupture beneath a site.
Aging is unforgiving. Reactor pressure vessels embrittle after decades of neutron bombardment, raising risk under stress. Past incidents, including Maanshan’s 2001 “full blackout” that cut both external and emergency power for two hours, highlight systemic vulnerabilities that do not improve with time.
Beneath the security and cost debate is the toughest issue of all: radioactive waste. After nearly 50 years of nuclear generation, Taiwan has more than 21,500 spent-fuel assemblies and over 210,000 drums of low-level waste — with no final repository for either.
High-level waste sits in crowded spent-fuel pools at closed plants designed for short-term cooling, not long-term storage. At Nuclear II, the pool grew so full that one reactor was shut early in 2021. Moving cooled fuel to interim dry-cask storage has been slowed by years of political gridlock and local opposition; the first dry-cask facility, at Nuclear I, began operating only in 2025 after long delays.
The official plan calls for a deep geologic repository by 2055. But enabling legislation has not passed, and site selection faces dense population, active geology, and entrenched “not in my backyard” opposition. A 2022 Control Yuan report underscored how difficult it will be to find a geologically suitable site in the country.
Low-level waste tells a parallel story on Orchid Island (Lanyu), home to the Tao (Yami) people. A “temporary” facility established in 1982 still holds roughly 100,000 barrels and remains a source of resentment and protest. Efforts to identify a permanent site have stalled for more than a decade. In 2012, the Ministry of Economic Affairs named Daren Township in Taitung and Wuqiu Township in Kinmen as candidates, but local governments refused to hold required referendums amid intense public opposition.
Opponents of a restart pose a blunt question. Is it responsible to generate more high-level waste by restarting 40-year-old plants when there is no scientifically viable or socially acceptable plan for the hazardous legacy already in storage?
Taiwan is caught between a reliable, low-carbon source that could support industry and crisis resilience, and unresolved questions about seismic risk, long-term costs, and intergenerational waste. The debate is not a simple pro- or anti-nuclear binary — it is a set of trade-offs with no painless option.
A credible path forward, regardless of nuclear’s ultimate role, cannot rest on political maneuver or slogans. It requires institutional trust and unusual transparency.
Any restart or new build should include a full lifecycle cost analysis — not just per-kWh figures — covering safety upgrades, decommissioning, and a funded, realistic waste plan.
Regulators must be empowered and resourced to make science-based decisions independent of political pressure, with public reasoning and international best practices as the standard.
Most of all, the government needs tangible progress on interim storage and an honest plan for final disposal. That means passing the legislation required to start repository siting, engaging potential host communities in good faith, and acknowledging that waste may be the most difficult domestic policy challenge Taiwan faces.
Shutting Maanshan was not an end but an inflection point. Taiwan’s energy future will be defined by how it confronts these trade-offs — with scientific integrity, democratic legitimacy, and a sense of duty to the generations that will inherit the consequences.
