A geologist who has spent more than four decades mapping Taiwan's faults is challenging Taiwan President Lai Ching-te's (賴清德) administration to back up its promise that a shuttered nuclear plant is "absolutely safe" with hard engineering evidence, not political assurance, before allowing it to restart.
Chen Wen-shan (陳文山), an emeritus professor of geosciences at National Taiwan University who has researched Taiwan's fault systems for more than 40 years, told Storm Media's New News magazine that his concern about Nuclear Power Plant No. 3, commonly known as the Maanshan Nuclear Power Plant, has nothing to do with being pro- or anti-nuclear. It comes down to a narrower physical question: the plant sits only about one kilometer from the Hengchun Fault, which has already been classified as active, and Chen wants to know what evidence the government will offer the public to prove the plant is safe. Chen has built his reputation studying active faults across Taiwan, including the Chelungpu Fault near the epicenter of the 1999 Chi-Chi earthquake, and is known in academic circles as the "father of fault dating" for his work reconstructing fault histories through carbon-14 analysis.
The Fault Was Mapped Decades Before Construction
Chen said the Maanshan plant's biggest inherent geological risk has always been the Hengchun Fault, which is not a recent discovery. Japanese geologists charted it on maps as far back as Taiwan's colonial period. When the plant's site was assessed decades ago, surveyors simply lacked sufficient evidence at the time to establish that the fault was active.
"The conclusion back then was that there was no data proving it was an active fault," Chen said. He argued that statement contains a basic logical flaw: the absence of evidence that a fault is active is not the same as evidence that it is not.
Taiwan's active-fault surveying, paleoseismology and dating methods were all far less developed when the plant was built, Chen said, and the government of the time had already committed to construction, bringing in American experts to review the site. Without firm proof that the Hengchun Fault was active, the plant went ahead on schedule as a political decision. Decades later, Chen said, the science has moved on: subsequent research has shown the fault has been active and that the surrounding rock has been fractured by movement along it. Under the current classification system used by Taiwan's geological survey authorities, older tiers that separated faults by certainty of activity have been dropped, leaving the Hengchun Fault simply classified as active. "An active fault is a fact. The geological risk is a fact," Chen said.
Chen Says the Fault's Long Silence Is Not Reassuring
Some may ask why, if the Hengchun Fault is so significant, it has not produced a major, well-remembered earthquake in recent history. Chen's answer runs the opposite direction. "That's exactly what makes it worse," he said.
A long quiet stretch tells geologists nothing about future safety, Chen explained. What matters is a fault's activity cycle, the timing of past major ruptures and how much stress has built up in the meantime. The trouble, he said, is that long-term paleoseismic records and monitoring data for the Hengchun Fault remain thin. He pointed to Japan, where paleoseismic studies in some regions have reconstructed major earthquakes stretching back one to two thousand years, not from modern instrument records but through painstaking geological fieldwork. "Many of the world's major earthquakes have no historical record at all," Chen said, arguing that a recent lack of major shaking cannot by itself justify confidence in the plant's safety, particularly since a nuclear plant is not an ordinary structure. Nuclear facilities carry extremely low-probability risks that can produce extremely high losses if they materialize, he said, so safety standards cannot rest only on decades without an incident; they must answer whether the plant would hold up under the most extreme scenario.
Aging Pipes Worry Chen More Than Reactor Collapse
Chen said discussions of Nuclear Power Plant No. 3's restart tend to overlook another point. Most people picturing an earthquake at a nuclear plant imagine the reactor building itself toppling. Chen thinks the more realistic danger lies elsewhere, in the plant's vast and complex network of internal piping and equipment.
"I'm not saying the earthquake would knock the plant down. I'm saying the aging pipes inside could come loose or buckle. That's where the real danger is," he said. Running a nuclear plant depends on cooling water, steam, emergency cooling and a web of safety systems, and strong shaking can cause pipe joints, supports or equipment to fail even if the building stands. For a plant that has operated for decades to extend its service life, Chen said, genuine seismic reinforcement cannot stop at the structure. It has to reexamine a huge volume of internal systems.
That is why Chen believes the word "reinforcement" is easy to say but enormous to execute. A retrofit with real substance would require shutting the plant down to inspect, and in many cases replace, aging pipes and related equipment throughout. "Real reinforcement means checking every pipe inside and replacing whatever needs replacing," he said, adding that such work cannot be finished in a few months. Drawing on Japan's experience with similar projects, Chen said a genuinely comprehensive retrofit could take years, at least two of them, meaning a government pushing for a fast restart could find its timeline colliding with the safety work itself.
A Safety Pledge Needs Evidence Behind It
Lai's administration has made "absolute safety" a central condition for restarting nuclear power, and Chen said that is precisely where the problem lies. A guarantee of safety cannot simply be a promise from a politician, he argued; it has to be translated into engineering standards and scientific evidence that can actually be checked.
The plant's roughly one-kilometer distance from the active Hengchun Fault will not disappear because energy policy changes, Chen said, and an active fault will not become inactive just because it gets reinvestigated for two or three years. He argued that the real question for the government is not whether one supports or opposes nuclear power, but a set of concrete engineering questions: What seismic assessments has the plant actually completed? Have its aging pipes and critical equipment been fully inspected and replaced where necessary? Is the plant's design basis sufficient to withstand the largest earthquake the Hengchun Fault could produce? And what specific engineering work and scientific evidence underpins the government's claim of "absolute safety"?
If the plant's structure or equipment falls short, Chen said, it should be reinforced to a standard that can genuinely withstand an earthquake. "You're making a promise to the public. You need to tell people exactly what that promise is based on," he said, calling it the central issue in the current restart debate.
Taipower's Survey Will Not Erase Fault Data
Taiwan Power Co., the state-run utility commonly known as Taipower, has already begun a fresh geological investigation of the site, and Chen is participating as a committee member, though he is not personally conducting the survey. Even so, he said he has little doubt about what a new investigation will show. "A survey can't run for two, three, or four, five years and turn an active fault into an inactive one," he said.
Further study can sharpen scientists' understanding of a fault's exact location, its activity timeline, and the likely magnitude and recurrence of associated earthquakes, Chen said, but existing geological evidence will not vanish simply because energy policy needs it to. He argued that if the government wants a new survey to settle whether the plant can restart, the harder and more relevant question is not simply whether the Hengchun Fault is active, but whether the plant can withstand the earthquake risk that an active fault next door creates. The two questions are entirely different, he said: one is a matter of geology, the other involves seismic engineering, nuclear safety, equipment aging and risk management.
Chen Says Restart Is Possible After Real Reinforcement
Chen has not reduced his position to a flat "the plant can never restart." His stance is closer to a conditional argument: if the government genuinely believes Taiwan's energy supply requires Nuclear Power Plant No. 3, it must first complete safety engineering thorough enough to convince the public. "If Lai Ching-te wants to restart the plant, that's not out of the question either. Then go ahead and properly reinforce it," Chen said, adding that he has not yet seen the government show the corresponding engineering resolve.
If large-scale reinforcement is genuinely needed, Chen said, the government should carry it out openly. Even opponents of nuclear power would still object, he acknowledged, but at least officials could point to completed engineering work and tell the public: "We've already done this much. Please trust that the plant is safe." That, he said, is fundamentally different from simply asserting the plant is safe. Chen believes the government is caught in a political bind, wanting to preserve the option of restarting the plant while worrying that a visible, large-scale reinforcement campaign would be read as a decision to fully return to nuclear power, leaving it without a clear public position. Energy policy can stay vague, Chen said. Geological conditions cannot.
Energy Policy Bends While Geology Stays Fixed
In mid-July, several business leaders met with President Lai at the presidential residence to discuss energy, water resources and industrial development, with corporate anxiety over surging AI-era electricity demand helping push nuclear restart back onto the policy agenda. Chen does not dispute that Taiwan faces a real power supply challenge, but he argues that shortage has to be addressed through a mix of energy policies rather than by sidestepping a nuclear plant's own geological and engineering safety questions.
Energy policy is a choice, he said, but geology cannot be negotiated. For a researcher who has studied faults for more than 40 years, whether the Hengchun Fault is active is not a partisan question and not a proxy fight between nuclear supporters and opponents, but a geological question already backed by substantial evidence. The genuinely political decision, in Chen's view, lies elsewhere: knowing an active fault sits beside the plant, how much risk is Taiwan willing to accept, and how much time and money is the government prepared to spend bringing that risk down to a level society can live with.
Chen noted that his mother was from Pingtung County, making him, as he put it, "half a Pingtung native." Pingtung is not short of land, he said, and if Taiwan genuinely views nuclear power as an indispensable part of its future energy mix, the more useful question may not be how quickly to restart a decades-old plant sitting beside an active fault, but what standard Taiwan should apply when rethinking the safety and siting of nuclear facilities altogether.
When the plant was built, people could say there was no evidence the Hengchun Fault was active, Chen said. Four or five decades on, that statement no longer holds. Technology advances, energy policy shifts and political parties can even change position, he said, but the fault underground does not move because of political convenience. In the end, Chen said the debate over restarting Nuclear Power Plant No. 3 comes down to a single question for the Lai administration: if the government says it will only restart the plant once it is "absolutely safe," what evidence is it prepared to produce to prove that, given an active fault sitting about one kilometer away?
