U.S. Gen III+ small modular reactor construction totals $34.9 billion of cumulative capital spending over 2026–2035 in the base case, and on the record so far who absorbs first-of-a-kind cost-overrun risk, not licensing tempo, gates when construction starts.

On 28 September 2026 the NRC issued its first construction permit for a commercial light-water small modular reactor, to TVA’s BWRX-300 at Clinch River in Tennessee, four months ahead of schedule. This report sizes U.S. Gen III+ SMR construction at $34.9 billion of cumulative capital spending over 2026–2035 in the base case, with a 50% interval of $21.8–45.8 billion.

The report’s thesis is that the first U.S. light-water SMR construction permit has shifted the binding constraint on deployment from the regulator to the balance sheet. On the record so far, who absorbs first-of-a-kind cost-overrun risk, not licensing tempo, gates when construction starts, while forging and labour capacity set how long it takes.

The 60-page report includes 12 charts and figures, 9 data tables, a forecast to 2035 under three named scenarios and a probability distribution, and profiles of 10 companies, including GE Vernova Hitachi Nuclear Energy, Tennessee Valley Authority, Holtec International, NuScale Power and Blue Energy.

Report Highlights:

• Built Where Overrun Risk Is Cheapest to Place: On the record so far, the first U.S. light-water SMRs will be built where first-of-a-kind overrun risk is cheapest to place, not in the order they were licensed. Federal cost-share can never exceed half of a project’s cost. Five projects carry first-unit risk with five different instruments, from federal match plus the owner’s own power revenue to vendor-paid milestones under a non-binding buyer. None has yet closed binding offtake, construction financing and a construction date.
• Cumulative Construction Capital, Read Correctly: U.S. Gen III+ SMR construction capital totals $34.9 billion over 2026–2035 in the base case, cumulative, not annual (50% interval $21.8–45.8 billion). The report also gives the probability-weighted view across its full distribution. Annual spending rises from $1.52 billion in 2028 to $5.16 billion in 2035. That is a 19.1% CAGR for a first-wave build-out that peaks in 2034 and falls in 2035, not steady market growth.
• The Largest Allocation-Model Segment: On the report’s five-unit reading of Blue Energy, venture-backed merchant development is the largest of the four allocation-model segments the report sizes. It rests on about five BWRX-300 units at Blue Energy (some 1.5 GW, as reported), whose final investment decision is still ahead, in 2027. First units at their sites carry more than half of base-case spending: the share that bears the first-of-a-kind premium and the allocation question. The report sizes each segment and the first-unit share.
• Vendors Financing Their Own Buyers: A vendor has begun paying a company its own filing calls a prospective customer. NuScale expensed $507 million for Milestone Contribution 1 to its developer partner ENTRA1 on the strength of a non-binding agreement with TVA for up to 6 gigawatts. It paid $247.5 million in cash in 2025 and the remaining $259.9 million in January 2026. At Palisades, Holtec is the technology vendor, the plant constructor with Hyundai Engineering & Construction, the operator and the merchant seller of the power.
• First Power Against the Realised Record: Under the report’s scenario probabilities, it is more likely than not that no U.S. Gen III+ unit reaches first power by 2035; the report gives the probability and how sensitive it is to the weights. The realised first-of-a-kind and first-in-country light-water builds located took 8.2 to 18 years from nuclear construction start to commercial operation, grid connection or commissioning. The three located Western cost outcomes ran 2.1 to 4.0 times their initial estimates. That timing covers light-water designs only: the publisher’s February 2026 report, which also counts non-light-water SMRs, has initial SMR demonstrations reaching commercial operation around 2032.

This report will provide answers to the following questions:

• How large is U.S. Gen III+ small modular reactor construction spending over 2026–2035, and how does it range across three named scenarios and a 27-scenario probability distribution?
• Who carries first-of-a-kind cost-overrun risk at each project on the U.S. roster, and which allocation event does each construction start depend on?
• What do the federal cost-share terms and milestones require, and what remains open at Clinch River after its construction permit?
• How is construction spending split by allocation model, by reactor design and by unit position?
• How long did realised first-of-a-kind and first-in-country light-water builds take, and how did their costs compare with their initial estimates?
• How likely is it that a U.S. Gen III+ unit reaches first power by 2035, and how sensitive is that probability to the report’s weights?
• Which forging, labour and supply-chain constraints set how long construction takes?

Companies Covered:

Tennessee Valley Authority, GE Vernova Hitachi Nuclear Energy, Holtec International, NuScale Power, ENTRA1 Energy, Blue Energy, Indiana Michigan Power, Duke Energy, Ontario Power Generation, BWX Technologies

Report Specifications:

60 pages · 12 charts and figures · 9 data tables · 10 company profiles · forecast horizon: 10 years, to 2035 · 3 named scenarios (conservative, base, optimistic) within a 27-scenario probability distribution · 3 segmentation views

Pricing:

• Single User License – $4,950
• Site License – $5,450
• Enterprise License – $5,950

To place an order, contact david.pringmill@policy2050.com