Nuclear Power Economics

Nuclear power is the most capital-intensive energy source: a single plant costs $10–20 billion, takes 10+ years to build, and must run for 40+ years to recover investment. Yet once operational, it produces baseload electricity at near-zero marginal cost and zero carbon emissions. Nuclear power economics is fundamentally a bet on long-term cost amortization, government backing, and stable regulatory environments.

The capital cost trap and first-of-a-kind premiums

Building a nuclear plant is arguably the most financially complex infrastructure project. First-of-a-kind plants in a region cost 50–100% more than subsequent builds, because regulatory approval, design validation, and supply-chain establishment are expensive. The Vogtle expansion (Units 3 & 4 in Georgia, completed 2024) cost over $35 billion for two units—roughly $17.5B per unit. Comparable coal plants cost $2–3B per unit; modern natural gas plants cost $0.5–1B.

This capital intensity creates a tragic mismatch: nuclear plants require upfront government subsidies (loan guarantees, construction support) because no private investor can finance $20 billion over a 10-year build with regulatory uncertainty. Yet once built, the plant generates cash for 60+ years, making it enormously valuable. The public finances the risk; private operators capture the reward.

Levelized cost of electricity: the deferred-cost illusion

The levelized cost of energy (LCOE) spreads the capital cost over the plant’s lifetime. A $10B plant producing 1 GW of power at 90% capacity factor will generate ~800 gigawatt-hours per year. Over 60 years, it generates 48,000 GWh. Dividing the $10B capital cost (plus operating costs of ~$50M/year) across 48,000 GWh yields an LCOE of ~$80–100/MWh.

This looks competitive with solar ($50–70/MWh) and wind ($40–60/MWh) on paper. But the comparison obscures a critical reality: the nuclear plant requires $10 billion upfront; solar and wind require 20% of that, spread across 20+ installations. The real constraint is capital availability, not unit economics. Few investors will commit $10B on a single industrial project.

Baseload value and capacity factor reliability

The economic advantage of nuclear is baseload reliability. A nuclear plant runs 90%+ capacity factor in steady state—it produces power around the clock, rain or shine. Solar produces only during daylight (20–25% capacity factor); wind only when windy (35–40%). This means:

• To get the same total energy output, you need 4x more solar capacity or 2.5x more wind capacity than nuclear
• Baseload plants (nuclear, coal, natural gas with high utilization) can replace each other; renewables cannot fully replace baseload without massive battery storage (energy storage)
• A grid with high renewables penetration needs either nuclear baseload or massive storage investment (currently uneconomical at scale)

This is why some grid operators and climate advocates argue nuclear is essential for decarbonization—it is the only carbon-free baseload source currently available.

Government subsidies and the economics of impossibility

No nuclear plant built in the US in the last 40 years has been financed without government support. This is not accident; it is necessity. Vogtle 3&4 received federal loan guarantees (reducing borrowing costs), construction tax credits, and favorable power purchase agreements with Georgia utilities. The 2023 Inflation Reduction Act extended production tax credits for nuclear plants.

Without these supports, nuclear plants cannot achieve return on equity acceptable to investors. The capital is too large, the construction timeline too uncertain, and the regulatory environment too volatile. Yet with subsidies, nuclear becomes competitive—and the public bears the financial risk.

This creates a political bind: nuclear is necessary for decarbonization, but politically difficult to finance because the costs are so visible and upfront. In contrast, subsidies for renewables (often delivered through tax credits or investment tax incentives) are less obvious to the public.

Decommissioning and tail-end liability

Once a nuclear plant stops generating (typically at 60 years), the reactor must be decommissioned—the radioactive core removed, the site decontaminated, and radioactive waste sequestered. Decommissioning cost $1–2 billion per plant. By law, utilities must establish a reserve fund to pay for this. The fund earns investment returns and grows over the plant’s lifetime.

In principle, the plant’s profits pay for its own decommissioning. In practice, plants are often sold to third-party operators, and disputes arise over who bears the decommissioning liability. Some utilities face shortfalls. Federal law requires decommissioning completion, but funding is sometimes inadequate.

Climate economics and the missing carbon price

The fundamental problem with nuclear economics is that carbon emissions are not priced into competing fuels. A natural gas plant emits ~400 tons of CO2 per gigawatt-hour; this pollution is free (no carbon tax). The natural gas plant looks cheaper than nuclear purely on operating cost.

If carbon were priced at $50–100 per ton (as some climate analyses suggest it should be), natural gas would cost 2–4x more. Nuclear would become far more economically attractive. However, without a carbon tax or binding emissions cap, nuclear must compete against subsidized fossil fuels and increasingly cheap renewables. Under these conditions, new nuclear is economically difficult (though existing nuclear is valuable because it has no fuel cost).

The case for advanced reactors and SMRs

Some analysts argue small modular reactors (SMRs)—plants under 300 MW, potentially mass-produced—could solve nuclear economics by:

• Reducing capital per unit (shorter construction, manufacturing scale)
• Allowing mass production and cost reduction
• Fitting sites unsuitable for large plants
• Providing heat for industrial processes, not just electricity

However, SMRs are currently far from economical: prototype costs run $3–5B per 300-MW unit, worse than large reactors. Mass production might reduce costs, but this requires first-of-a-kind investment that few private investors will finance.