Nuclear power produces almost no carbon dioxide while running, and it also produces waste that stays dangerous for millennia. Both halves of that sentence are true, which is exactly the kind of contention Truza exists to lay out. The question people actually argue about is narrower than the shouting suggests: is the spent fuel a solved engineering problem, an unsolved political one, or a lurking catastrophe? Here is what the primary sources on each side actually claim, and how strong their evidence is.
What the waste actually is
Spent fuel from a light-water reactor is about 95 percent uranium that simply was not consumed, roughly 1 percent plutonium, and the remainder is a mix of fission products, strontium-90, cesium-137, and transuranic elements. The fission products are what make fresh spent fuel ferociously radioactive: a bundle removed from a reactor core emits enough decay heat to require continuous cooling in a water pool for several years. After about five to ten years, radioactivity has fallen enough for the fuel to be moved to dry casks, steel and concrete containers that passively shed the remaining heat. This much is not disputed by anyone; it is standard nuclear engineering, documented in IAEA and national regulator technical reports.
The dispute starts at the next question: what happens after dry casks, and for how long. The isotope that sets the timescale is not the famous plutonium-239 (half-life 24,100 years) but the longer-lived transuranics beyond it, which is where figures like “dangerous for 100,000 years” or “a million years” come from. Different numbers come from different assumptions about what dose counts as hazardous, which is why the argument never resolves by quoting a number at the other side.
The case that it is a solved engineering problem
The strongest version of the pro-nuclear position rests on three claims. First, dry cask storage has an actual decades-long operational record: the United States has stored spent fuel in casks since the 1980s with no radiation release to the public, and Finland has built a permanent repository, Onkalo, designed to hold the fuel for a hundred thousand years. Onkalo is not a concept; its tunnels are excavated in bedrock and its approval process is complete. Second, the volume is small. All the spent fuel ever produced in the US would fit on a single football field stacked about ten metres high, and a gigawatt-year of reactor operation produces on the order of a few tens of tonnes of spent fuel, against millions of tonnes of ash and CO2 for the coal equivalent. Third, the waste is contained and accounted for, which is more than can be said for combustion waste that is simply dumped into the atmosphere.
The case that it is unsolved
The strongest opposing case is political and economic rather than physical. No country except Finland has actually opened a permanent deep geological repository for civilian spent fuel. Sweden approved one at Forsmark only in 2022. The US designated Yucca Mountain in 1987, spent billions, and never opened it; the project was defunded and the site remains a legal ghost. That is a record of one success against decades of failure, and the failure mode is not engineering but trust: communities do not want to be the site, and the timescales exceed the life of any institution that promises to guard the material.
The second prong is economics. In several countries the fee charged per kilowatt-hour to pay for disposal was based on projected costs that keep being revised, and in the US the nuclear waste fund collected tens of billions of dollars while disposing of nothing. Critics argue this makes “the waste problem is solved” a claim about physics that quietly assumes away the politics and the balance sheet.
Reprocessing, and why countries disagree
France reprocesses spent fuel at La Hague, chemically recovering plutonium and uranium to make mixed-oxide fuel for another cycle, which reduces the volume of high-level waste and extracts more energy. The United Kingdom historically did the same; the United States abandoned reprocessing in the 1970s largely over proliferation concerns, since separated plutonium is weapons material. Whether reprocessing counts as “recycling the waste” or “creating a proliferation risk and a plutonium stockpile” is itself a live dispute, and the positions track almost perfectly with each country’s institutional history.
Where the two sides genuinely talk past each other
The engineering side says: the material is inert inside a cask, the dose at the fence is lower than a cross-country flight, and a repository in stable bedrock is provably safe for the relevant timescales. The sceptical side says: the casks are temporary by design, “temporary” has already meant seventy years in some countries, and the long-term safety case rests on projections about geology and institutions that cannot be tested in advance. Both are factually defensible. The disagreement is about how much weight to give an unopened promise versus an unbuilt promise, which is a judgement about institutions, not about radiation.
Truza’s summary of the rating landscape
Technical claims about cask integrity, isotope half-lives, and waste volumes are supported by peer-reviewed and regulatory sources (four to five stars under our ranking system, and the two sides rarely contradict each other on these). Claims that disposal “is solved” or “can never be solved” are opinion-laden extrapolations; they typically rate three stars or receive a context note rather than a rating, because no source can prove an institutional outcome a hundred thousand years out. A comparable case of trial-data-versus-extrapolation that we have already laid out side by side is statins for primary prevention.
If you have a better source than the ones behind our nuclear waste entries, submit it. Truza improves one citation at a time.