Nuclear Energy: A Vital Tool for a Clean Future or an Unacceptable Environmental Gamble?

Nuclear energy is essential not because it is risk-free, but because climate policy must be judged against realistic alternatives. Nuclear plants deliver large amounts of low-carbon electricity around the clock, complementing wind and solar when weather, seasons, or limited storage constrain their output. Countries that retire reactors prematurely often prolong fossil-fuel use, undermining the central goal of rapid decarbonization.

The major objections are serious but manageable. Modern reactor designs incorporate passive safety systems and lessons from Chernobyl and Fukushima, while nuclear power’s overall deaths per unit of electricity remain far below those caused by coal, oil, and air pollution. Radioactive waste is small in volume, carefully contained, traceable, and technically suitable for deep geological disposal; unlike fossil-fuel emissions, it is not routinely dispersed into the atmosphere. Cost and construction delays demand better regulation, standardized designs, and disciplined project management—not abandonment of a proven low-carbon technology.

Renewables should expand rapidly, but relying on them alone would require immense transmission, storage, overbuilding, and backup capacity. A resilient clean grid should combine renewables, nuclear power, storage, efficiency, and other low-carbon resources. Excluding nuclear narrows our options precisely when speed, reliability, and scale matter most. The responsible course is therefore to retain safe existing reactors and expand modern nuclear power alongside renewables, rather than gamble the climate on a single technological pathway.

Nuclear power fails the test that matters most in a climate emergency: delivering the maximum amount of clean electricity, as fast as possible, at the lowest cost and risk. On every one of those measures, it loses.

First, cost and speed. Climate change is a race against the clock, and nuclear is the slowest, most expensive horse in the stable. Recent Western projects tell the story: Vogtle in Georgia came in at over 30 billion dollars, roughly double its budget and seven years late. Hinkley Point C in the UK and Flamanville in France have followed the same pattern of decade-long delays and cost explosions. Meanwhile, the cost of solar has fallen by roughly 90 percent and wind by about 70 percent over the past decade, and a utility-scale solar or wind farm can be built in one to three years. Every dollar spent on nuclear is a dollar that could have cut two to five times more carbon, sooner, through renewables, storage, grid upgrades, and efficiency. That is not ideology; that is arithmetic. Capital is finite, and misallocating it delays decarbonization.

Second, catastrophic risk. Defenders of nuclear speak of low probability. But the defining feature of nuclear risk is not its frequency; it is that its consequences are effectively unbounded and irreversible. Chernobyl rendered a region uninhabitable for generations. Fukushima forced the evacuation of about 150,000 people and has cost Japan hundreds of billions of dollars, with contaminated water still being discharged more than a decade later. Both plants were operated by advanced industrial states with regulators who insisted they were safe. No solar panel has ever required a permanent exclusion zone. And in an era of extreme weather, cyberattacks, and armed conflict, we have watched the Zaporizhzhia plant become a hostage in a war zone. You cannot weaponize a wind turbine.

Third, waste. After seventy years of commercial nuclear power, not one country on Earth operates a permanent high-level waste repository at scale. Finland's Onkalo is the sole facility approaching operation; the United States abandoned Yucca Mountain after decades and billions of dollars, leaving roughly 90,000 tonnes of spent fuel sitting in temporary pools and casks at more than seventy sites. Asking future generations to guard material that stays hazardous for tens of thousands of years, when no institution in human history has lasted that long, is not stewardship. It is deferral dressed up as engineering.

Fourth, the reliability argument is outdated. Baseload inflexibility is a liability, not a virtue, on a modern grid. The real solution to intermittency already exists and is scaling fast: geographic diversification, transmission interconnection, batteries, pumped hydro, green hydrogen, demand response, and firm renewables like geothermal and hydro. Battery storage costs have collapsed, and grids in South Australia, Denmark, and California now run for extended periods on very high renewable shares. Nuclear plants, by contrast, must be shut down in heatwaves and droughts when cooling water is too warm or too scarce, precisely when demand for power peaks. France has repeatedly lost large portions of its nuclear fleet to corrosion issues and river temperatures.

Finally, small modular reactors are a promise, not a product. NuScale's flagship US project was cancelled in 2023 after costs rose about 50 percent before a single unit was built. Betting the climate on a technology that does not yet exist commercially, while proven alternatives are cheap and deployable today, is a gamble we cannot justify.

We do not need to accept an unbounded, multi-generational hazard and a decades-long construction timeline. We need to build what works now.

The opposition’s case relies on comparing the worst nuclear projects with the cheapest individual renewable generators, rather than comparing complete electricity systems. Solar and wind are indispensable, but their headline costs exclude much of the transmission, storage, overbuilding, curtailment, and firm backup needed to provide electricity through nights, wind droughts, and seasonal shortages. Batteries are excellent for shifting power by hours; affordable storage across days or seasons remains limited. High-renewable regions cited by the opposition still depend on interconnectors, dispatchable generation, or neighboring grids. A clean system needs dependable low-carbon capacity, not merely cheap energy when weather permits.

Vogtle, Flamanville, and Hinkley demonstrate failures of financing, regulation, and one-off construction—not a physical law that nuclear must be slow or prohibitively expensive. Standardized programs in France, South Korea, and elsewhere have historically built reactors much faster. More importantly, safely extending existing plants is often among the cheapest and quickest ways to preserve low-carbon electricity. Closing them while fossil generation remains is climate self-sabotage. We need not bet on small modular reactors to justify established reactor technology.

Calling nuclear consequences “unbounded” is rhetoric, not risk analysis. Chernobyl used a deeply flawed reactor design without a full containment structure, while Fukushima’s direct radiation death toll was far smaller than the disruption caused by a poorly managed evacuation. Every energy system has hazards, and the fair comparison is total harm per unit of electricity. Nuclear performs among the safest sources by that measure, whereas fossil fuels kill through routine air pollution and climate damage. Modern reactors add passive shutdown, stronger containment, and more robust cooling. War and cyber threats demand strict siting and security standards, but they do not justify discarding an entire low-carbon technology.

Waste is also a governance challenge with a demonstrated engineering solution, not an unsolved scientific mystery. Spent fuel is small in volume, contained, monitored, and can be isolated in stable geological formations. Finland’s Onkalo does not prove failure; it proves that permanent disposal is achievable when political institutions follow through. Unlike fossil waste, radioactive material is not continuously released into the shared atmosphere, and its hazard declines over time.

Finally, nuclear need not crowd out renewables. A rational strategy expands wind, solar, grids, storage, efficiency, geothermal, and nuclear according to local conditions. The truly dangerous gamble is to prohibit a proven source that already supplies enormous quantities of low-carbon electricity and assume that every future gap in an overwhelmingly weather-dependent system will be filled cheaply and on time. Climate urgency calls for more viable tools, not fewer.

My opponent's case rests on four claims. Each collapses under scrutiny.

First, the claim that nuclear is judged unfairly against "realistic alternatives." But realism is precisely my argument. The realistic question facing a utility or a finance minister is not "nuclear versus nothing" — it is "where does the next hundred billion dollars of clean-energy capital go, and how much carbon does it displace by 2035?" My opponent never answers this. Nuclear's own advocates concede a 10-to-17-year timeline from decision to grid connection in Western democracies. Carbon emitted in the 2020s and 2030s is what determines whether we cross irreversible tipping points. A technology that delivers its first electron in the late 2030s is not a climate solution for this decade; it is a bet that we can afford to wait. We cannot.

Second, the deaths-per-terawatt-hour comparison. Comparing nuclear to coal is a rhetorical sleight of hand, because coal is not on the table. Nobody in this debate is proposing coal. Compare nuclear to what it actually competes with — solar, wind, geothermal, storage — and its safety advantage vanishes entirely, while its cost disadvantage becomes stark. And that statistic measures only direct mortality. It does not price the permanent loss of territory, the destruction of communities, the collapse of property values, or the hundreds of billions in cleanup liabilities. Those costs are real, and in every country they are ultimately socialized. It is worth noting that no private insurer on Earth will fully underwrite a nuclear plant against catastrophic accident. Governments must cap liability by statute for these plants to be financeable at all. When the insurance industry — an industry that exists purely to price risk — declines to price this one, that is not fearmongering. That is the market's verdict.

Third, waste. My opponent calls it "small in volume, contained, traceable, and technically suitable for deep geological disposal." Notice the word "suitable." Not disposed of — suitable for disposal. After seventy years and roughly 400,000 tonnes of spent fuel worldwide, exactly one repository is nearing operation, in a single country with unusually stable bedrock and unusual political consensus. The United States spent decades and billions on Yucca Mountain and abandoned it. Germany's Asse II and Gorleben sites failed. "Technically viable" is a claim about physics; the obstacle has always been geology, politics, and institutional durability across timescales longer than any civilization has ever sustained. Calling that solved is an act of faith, not engineering.

Fourth, and most importantly, my opponent's central pillar — that renewables alone would require "immense transmission, storage, overbuilding, and backup." Yes, they would. And here is what he omits: every credible study finds that portfolio is still cheaper and faster than the nuclear-inclusive pathway. Transmission lines take three to six years. Batteries are deployed in months. Solar farms are built in eighteen months. These are all technologies on steep cost-decline curves — solar down roughly 90 percent, batteries down roughly 85 percent, wind down roughly 70 percent in a decade. Nuclear is the only major energy technology in history with a documented negative learning curve: it has become more expensive over time, in every country that has built it at scale, including France.

He also never rebutted my reliability point. Nuclear is not "always on." France lost more than half its fleet to corrosion inspections in 2022 and has repeatedly curtailed output when rivers ran too warm for cooling. In a warming world, thermal plants dependent on cool water are becoming less reliable, not more, and they fail precisely during the heatwaves when demand peaks. Meanwhile inflexible baseload actively obstructs a high-renewables grid, because it cannot ramp to follow variable supply. My opponent is defending a grid architecture from 1975.

Finally, the accusation that I would "gamble the climate on a single technological pathway." This inverts the truth. I am proposing a genuinely diverse portfolio — solar, onshore and offshore wind, geothermal, hydro, batteries, pumped storage, long-duration storage, demand response, efficiency, and interconnection. That is eight or more independent technologies with independent supply chains and independent risk profiles. My opponent wants to divert scarce capital and scarce engineering talent into a single technology with a fifty-year track record of overrunning its budget, and whose next generation — the small modular reactor — has yet to produce a single commercial kilowatt-hour. NuScale's flagship American project was cancelled in 2023 after costs rose roughly 50 percent before construction began. That is not a hedge. That is the concentrated bet.

There is one final asymmetry that no argument can dissolve. If a solar farm fails, you lose electricity. If a nuclear plant fails, you lose a region — for generations. Both risks are manageable in expectation. Only one is unrecoverable. When a cheaper, faster, safer path exists, accepting an unbounded and irreversible downside is not courage. It is a gamble we have no need to take.

The opposition offers a false choice: either build renewables now or build nuclear later. A serious climate strategy does both—rapidly deploy wind, solar, grids, storage, and efficiency while preserving existing reactors and developing dependable low-carbon capacity for the decades ahead. Climate change does not end in 2035, and electricity demand will grow through electrification, development, and data-intensive industries. Abandoning firm nuclear power merely because some projects have been badly managed risks locking in gas backup for generations.

The claim that every credible analysis finds renewable-only systems cheaper is far too sweeping. Results depend heavily on geography, transmission availability, weather assumptions, storage duration, permitting, and the cost assigned to firm capacity. Generator prices are not system prices. Batteries handle hourly balancing well, but prolonged wind and solar shortages remain a different challenge. Interconnection and demand response help, yet cannot create electricity during a widespread supply deficit. Nuclear does not need to run perfectly or inflexibly to add value; it needs to provide substantial, weather-independent low-carbon output, and reactors can load-follow where markets reward it. A temporary French fleet problem no more invalidates nuclear than a regional wind drought invalidates wind.

Nor has the opposition established that nuclear presents an “unbounded” danger. Chernobyl and Fukushima were grave events, but decades of evidence still place nuclear among the lowest-mortality energy sources per unit generated. Liability rules reflect the unusual scale and correlation of rare accidents, not proof that risk cannot be assessed; governments also underwrite floods, terrorism, financial crises, and other risks private insurers cannot absorb alone. Modern containment, passive safety, better regulation, and careful siting directly reduce nuclear risk. Meanwhile, fossil generation used to cover clean-power shortfalls causes routine deaths and climate damage rather than merely hypothetical harm.

Waste is difficult politically, but technically manageable. Onkalo matters precisely because it demonstrates that deep geological disposal can move from theory to implementation. Nuclear waste is compact, inventoried, contained, and declining in hazard. The fact that governments have delayed repositories does not make the underlying engineering impossible; it shows that durable policy is required. By contrast, carbon waste from fossil fuels is released at enormous scale into a shared atmosphere with no containment at all.

Most importantly, our position does not depend on unproven small modular reactors, nor does it demand nuclear everywhere. It supports extending safe existing plants, standardizing proven designs where conditions are suitable, improving construction discipline, and expanding nuclear alongside renewables. The opposition’s diverse list conspicuously excludes a proven technology already supplying vast amounts of low-carbon electricity and assumes that transmission, seasonal storage, geothermal, and other resources will all scale cheaply enough in every region. That is the more concentrated gamble: betting climate stability on the timely success of an exclusively non-nuclear system.

The responsible standard is not whether nuclear is flawless. No energy source is. It is whether retaining and expanding nuclear in appropriate places makes a clean grid more reliable, resilient, and achievable. Because it supplies firm low-carbon power while reducing dependence on fossil fuels, the answer is yes. Climate urgency requires managing nuclear’s risks—not discarding one of our strongest tools against a far larger planetary threat.

Let me close by returning to the only question that matters: with finite capital, finite time, and finite engineering talent, what actually cuts the most carbon fastest at the least risk?

My opponent never answered it. Not once across two turns did he name a single figure for what nuclear costs per tonne of carbon avoided, or a single date by which a new Western reactor ordered today would deliver power. Those omissions are the debate. Vogtle: over 30 billion dollars, double budget, seven years late. Hinkley Point C and Flamanville: the same story. Meanwhile solar is down roughly 90 percent, batteries roughly 85 percent, wind roughly 70 percent in a decade, and each can be built in months to three years. Nuclear remains the only major energy technology with a documented negative learning curve — it gets more expensive the more we build it, in every country including France. My opponent's answer was that these were failures of "financing, regulation, and one-off construction." But a technology whose costs depend on flawless institutions, perfect regulation, and uninterrupted political consensus across fifteen-year construction windows is not a reliable tool. It is a hope.

He asked us instead to admire France and South Korea in the 1970s and 80s. That concedes my point: nuclear's success cases come from a vanished era of state-directed construction, cheap capital, and weak public consultation. We must build in the world we actually inhabit.

On safety, he retreated to deaths per terawatt-hour versus coal. But coal is not on the ballot in this debate. Against the alternatives nuclear genuinely competes with — solar, wind, geothermal, storage — that safety advantage disappears entirely. And he never engaged my strongest evidence: no private insurer on Earth will fully underwrite a reactor against catastrophic accident. Governments must cap liability by statute or these plants cannot be financed. The one industry whose entire business is pricing risk refuses to price this one. That is not rhetoric; that is the market's verdict, and my opponent left it standing unchallenged.

He also called Fukushima's harm largely a product of a "poorly managed evacuation." Consider what that argument requires us to believe: that the technology is safe provided that governments, utilities, and emergency responders perform flawlessly under crisis conditions — the very assumption that failed at Chernobyl, failed at Fukushima, and is failing right now at Zaporizhzhia, where a reactor complex sits hostage inside an active war zone. You cannot hold a wind farm hostage.

On waste, his own words settled it. He said spent fuel is "technically suitable for deep geological disposal." Suitable — not disposed of. Seventy years, roughly 400,000 tonnes of spent fuel worldwide, and exactly one repository nearing operation, in a single country with unusually stable bedrock and unusual political consensus. Yucca Mountain: abandoned after decades and billions. Gorleben and Asse II: failed. He reframed this as a "governance challenge." But governance across tens of thousands of years is not a solved problem — no institution in recorded history has lasted even a fraction of that span. Calling that settled is faith, not engineering.

And he never rebutted reliability at all. France lost more than half its fleet to corrosion inspections in 2022 and has repeatedly curtailed output when rivers ran too warm to cool reactors. In a warming world, thermal plants fail precisely during the heatwaves when demand peaks. Inflexible baseload does not stabilise a high-renewables grid — it obstructs one, because it cannot ramp to follow variable supply.

Which leaves his final charge: that I am betting everything on one pathway. The reverse is true. I have proposed eight-plus technologies with independent supply chains and independent failure modes — solar, onshore and offshore wind, geothermal, hydro, batteries, pumped storage, long-duration storage, demand response, efficiency, interconnection. He proposes diverting scarce capital and scarce engineers into a single technology with a fifty-year record of overrunning budgets, whose next generation has yet to produce one commercial kilowatt-hour; NuScale's flagship American project died in 2023 after costs rose roughly 50 percent before construction started. That is the concentrated bet.

One asymmetry survives every argument made here tonight. When a solar farm fails, you lose electricity. When a reactor fails, you lose a region for generations — and the bill lands on the public, because no insurer will take it. Both risks are manageable in expectation. Only one is unrecoverable. When a cheaper, faster, genuinely diversified path is available and already scaling, accepting an unbounded and irreversible downside is not pragmatism. It is a gamble with no upside worth the stake — and we should decline it.