Is nuclear power cheaper than solar over a plant's lifetime?

On the most common way of measuring this — levelized cost of electricity, the total lifetime cost of building and running a plant divided by the total power it produces — the answer is no. New utility-scale solar is cheaper than new nuclear per megawatt-hour, and by a wide margin. On this all five independent assessments agree. Where they part ways is on whether levelized cost is the right frame for the question you're likely actually asking.

The axis of divergence is scope, not fact. Everyone accepts the same underlying economics; the disagreement is about what counts as a fair comparison.

The majority position takes the question at face value: compare the lifetime cost of a nuclear plant against the lifetime cost of a solar plant, per unit of energy delivered. On that basis solar wins clearly. Nuclear's cost is dominated by enormous upfront construction expense, long build times measured in a decade or more, financing charges that compound over those years, and a persistent tendency toward budget overruns. Solar has fallen dramatically in cost over the past fifteen years, installs quickly, and has almost no fuel or major operating cost once built. The lifetime arithmetic favors solar, and it isn't close.

The dissenting position accepts every one of those numbers but argues the plain comparison quietly smuggles in a false equivalence. Two refinements change the practical answer. First, new versus existing: the economics of building a nuclear plant today are punishing, but an already-built, paid-off nuclear plant is one of the cheapest sources of power on the grid, because its dominant cost is already sunk. If the real question is 'should we keep this reactor running,' the answer often reverses. Second, and more fundamental, a megawatt-hour of solar and a megawatt-hour of nuclear are not the same product. Nuclear delivers firm, dispatchable, round-the-clock power. Solar delivers intermittent power that vanishes at night and sags under cloud, and making it firm requires storage, overbuilding, or backup generation — costs that a bare levelized-cost figure for solar alone leaves out. Compare a system that must stay lit at 3 a.m. in January, and the gap narrows or, in some grid contexts, closes.

The crux is therefore a single question: does 'over a plant's lifetime' mean the raw cost of energy produced, or the cost of meeting actual, continuous demand? If the former, solar is decisively cheaper and the matter is settled. If the latter, the comparison depends on the specific grid, the availability of storage and its cost, how much firm capacity is needed, and whether you're weighing new construction or the extension of existing plants.

What would resolve it is precision about the question. For new-build, energy-only comparison: solar, unambiguously. For a system-level comparison that prices firmness and reliability, or for the decision to extend an existing reactor: the answer is genuinely contingent and can favor nuclear. No external data was consulted for this synthesis, so treat the direction of the finding as well-founded on established economics while regarding any specific cost figures as needing verification against current regional data — costs in this domain shift year to year and vary sharply by country and grid.