energy · family: the solution exists but nobody can afford it
no fleet withouta first buyer
Small modular reactors cannot achieve cost competitiveness without fleet production that requires a first customer willing to absorb uncompetitive costs
Problem statement
NuScale Power's VOYGR design — the first small modular reactor to receive NRC design certification, and at the time of the cancellation the only U.S. SMR design with NRC approval — had its first commercial project (the Carbon Free Power Project with UAMPS) canceled in November 2023 after estimated costs rose from $3 billion (for 600 MW, 12 modules, as announced in 2015) to $9.3 billion (for 462 MW, 6 modules, in 2023). The target electricity price rose from $55/MWh to $89/MWh; CATF describes UAMPS's market as one with cheap natural gas and growing wind deployment, and Utility Dive cites Lazard's April 2023 estimate that a newly built utility-scale solar facility with some storage produces power at an unsubsidized $46–102/MWh. The SMR business case rests on building many units — CATF's post-mortem calls on governments to incentivize "large order books" — but the first unit is the expensive one, and CATF reports that utilities, while interested in the technology, "hesitated to commit to the first SMR in the United States that had not even initiated the regulatory process." The CFPP was to have been, in E&E News's words, "the country's first expected commercial small modular reactor"; its cancellation left that first project unbuilt.
Why this matters
Proponents of SMRs argue that they will deliver cheaper carbon-free power that can run nearly around the clock, and CATF suggests that industries such as data centers and heavy industry, and public utilities with strategic goals and strong demand for zero-carbon reliable energy, may be better suited than investor-owned utilities to take on first-of-a-kind risk. Public money is already committed: DOE had provided $232 million to the CFPP since 2020 under a $1.4 billion cost-share deal. CATF counts more than 50 SMRs in the world today and says it is unrealistic to believe that all will succeed. Without a way through the first-of-a-kind trap, new nuclear may be limited to large conventional reactors, which have their own record of cost and schedule overruns: Utility Dive cites a 2014 study of 180 nuclear power projects in which 175 exceeded their initial budgets by an average of 117% and took on average 64% longer than projected, and Flamanville 3 in France, at an estimated 13.2 billion euros, four times the forecast at the start of construction.
What’s been tried and why it hasn’t worked
NuScale's design submerged the reactors in a large shared pool, which CATF says still requires significant on-site civil works and "incurred a large, fixed cost regardless of the number of modules" — a lack of modularity that made the design more expensive and less adaptable than other SMR alternatives. When subscriptions fell short of the level needed for the planned 12-pack, NuScale cut the project to a 6-pack and uprated each module from 50 MWe to 77 MWe to offset the rising marginal cost; CATF says that offset was insufficient for this design. The project was launched in 2015, before NuScale had applied for NRC design certification, because the company needed a customer to instill investor confidence; CATF says UAMPS, a collective of municipal utilities, lacked nuclear experience, could not take on cost risk on behalf of its customers, and operated in a market with cheap natural gas and growing wind. As costs overran and the operating date slipped from 2026, several member utilities withdrew from a 2019 agreement to buy 200 MW. DOE had provided $232 million since 2020, and NuScale said the project could not advance because it lacked enough subscribers. CATF's conclusion is that the CFPP was a demonstration project forced into a commercial model, and that the VOYGR — a relatively large and complex plant of up to 924 MWe — is less modular than the smaller designs other developers are pursuing; Ramana's opinion piece in Utility Dive argues the opposite lesson, that the loss of economies of scale means the cost trend "will afflict any small modular nuclear reactor project."
What would unlock progress
CATF's post-mortem proposes treating first-of-a-kind projects as demonstrations rather than full commercial endeavors: governments could make reactor sites available where developers deploy prototype or demonstration reactors through a streamlined process, with government offtake or insurance covering costs beyond a threshold, and with vendors paid on target milestones such as design completion or regulatory approvals rather than given a blank check. Governments could also incentivize large order books and cross-industry partnerships, on the model of DOE's Regional Clean Hydrogen Hubs program, to support the scale-up of the nuclear market. Customers better suited to first-mover risk — data centers, heavy industry, and public utilities with strategic zero-carbon goals — could replace investor-owned utilities as the first buyers. On the design side, CATF favors more modular and less complex SMR designs than the VOYGR, and on the regulatory side, risk-informed decision-making, ACRS reform, and a streamlined licensing process, since the traditional sequence of licensing a design and then building it is, in CATF's words, "too slow and expensive for today's dynamic market."
Entry points for student teams
A team could model the fleet learning curve for SMR deployment, identifying the minimum fleet size at which nth-of-a-kind costs become competitive with natural gas and renewables under different financing and regulatory scenarios. A policy-focused team could design a government procurement mechanism for first-of-a-kind SMR deployment based on military procurement precedents. Relevant disciplines: nuclear engineering, energy economics, public policy, systems engineering.
Genes
Structural cousins: stuck the same way, in other fields
Sources
Carlos Leipner, Jon-Michael Murray, Andrew Maxwell and Marcelo Botrel, CATF, "Lessons learned from the recently cancelled NuScale-UAMPS project," November 22, 2023, Zach Bright, E&E News by POLITICO, "NuScale cancels first-of-a-kind nuclear project as costs surge," November 9, 2023, M.V. Ramana, Utility Dive (opinion), "The collapse of NuScale's project should spell the end for small modular nuclear reactors," January 31, 2024, accessed 2026-02-20, re-read 2026-09-30 go to source 1 ↗ go to source 2 ↗ go to source 3 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
NuScale's CFPP cancellation was described by E&E News as a major setback to the advanced nuclear industry, and by CATF as a major though temporary setback, coming after NuScale had achieved the first ever design certification for an SMR. CATF counts more than 50 SMRs in the world today and says it is unrealistic to expect all to succeed; Ramana argues in Utility Dive that the cost trends seen at CFPP will afflict any SMR project. Distinct from energy-advanced-reactor-operations-cost (which covers ongoing O&M costs; this brief covers upfront capital costs and the fleet deployment paradox). The almost-cluster signal `cost-reduction` + `institutional-integration` + `economic` is directly relevant — resolving this problem requires coordinated cost-reduction across reactor vendors, construction firms, regulators, and power purchasers.
Reconciliation 2026-09-30: Fact repair after reading the whole write-up against every source it names. All three were fetched in full on 2026-09-30, directly from the publishers: the CATF post-mortem (https://www.catf.us/2023/11/lessons-learned-recently-cancelled-nuscale-uamps-project/, November 22, 2023), the E&E News by POLITICO report (https://www.eenews.net/articles/nuscale-cancels-first-of-a-kind-nuclear-project-as-costs-surge/, November 9, 2023) and the Utility Dive piece (https://www.utilitydive.com/news/nuscale-uamps-project-small-modular-reactor-ramanasmr-/705717/), which turned out to be an opinion column by M.V. Ramana published January 31, 2024, not 2023. (1) The Source line gave the CATF title without the word "recently" and dated the Utility Dive piece 2023; both titles and dates are now copied from the pages, with authors and links added for the two press sources. (2) The text said costs rose "from $3.6B (for 720 MW, 12 modules)". Utility Dive gives $3 billion for 600 MW and 12 modules as announced in 2015, then $4.2 billion in 2018 (when the plant became 720 MW), $6.1 billion in 2020 and $9.3 billion in 2023; the sentence now gives the 2015 and 2023 figures. (3) The text said the target price "surged 53%, from $58/MWh to $89/MWh". E&E News gives a target of $55 per megawatt-hour and a later figure of $89; no source gives $58 or a percentage, so the sentence now gives $55 and $89 and no percentage. (4) The text dated NRC design certification to January 2023 and said the cancellation came "just 10 months" after it. No source gives January 2023; E&E News says NuScale first got NRC approval for a 50-MW design in July 2022. The date and the ten-month interval were removed. (5) The text said the trap "has prevented any SMR from reaching commercial operation in the Western world". No source says this; the sentence now says what E&E News says, that the CFPP was to have been the country's first commercial SMR. (6) Removed from Why This Matters because no source contains them: nuclear's ~20% share of U.S. electricity, an average fleet age of ~42 years, retirements by 2050, "over $5 billion" invested in SMRs globally, hydrogen production as a use, and the AP1000 and EPR by name. The section now uses what the sources give: DOE's $232 million since 2020 under a $1.4 billion cost-share deal (E&E News), CATF's count of more than 50 SMRs, and Utility Dive's 2014 study of 180 projects and its Flamanville 3 figures. (7) The text said DOE provided "$230M"; E&E News says $232 million. (8) The parenthetical "seismic structures, containment, cooling systems" and the whole passage on "irreducible" safety-driven on-site costs and the "factory-built modular reactor" concept being misleading are not in any source; CATF ties the fixed cost to the large pool and calls the problem a lack of modularity in this design, and says the CFPP experience does not highlight a flaw in the SMR concept. The passage now follows CATF, and gives Ramana's contrary reading. The claim that renewables getting cheaper drove members to withdraw was softened to what E&E News says: cost overruns and delays. (9) The unlock section's microreactors under 20 MWe, TRISO fuel, passive decay heat removal, a "10+" unit fleet, the military nuclear vessel model, single-package factory modules, and "performance-based" licensing against "1970s" rules are in no source and were replaced by the measures CATF actually proposes: demonstration-first deployment on government sites with offtake or overrun insurance and milestone payments, large order books on the hydrogen-hubs model, first customers from data centers, heavy industry and public utilities, simpler and more modular designs, and risk-informed, streamlined licensing. (10) The notes above said the cancellation was "the most significant SMR setback to date" and named X-energy, Kairos and TerraPower; no source ranks the setback or names those companies. The sentence now gives the sources' own words (a "major setback"; a "major, though temporary, setback") and CATF's more-than-50-designs count. Confirmed in the sources and left as written: the November 2023 cancellation; the reduction from 12 modules to 6 and the 462 MW total; the shared pool as a fixed cost; UAMPS as a collective of municipal utilities; the relevance of cheap gas and wind; the entry points. Not checked: the spelling VOYGR (CATF writes "VOYAGR"; NuScale's own spelling was not looked up this round). Every quotation now in the write-up was checked against its source word for word. No title changed. Genome tags untouched. A second reader went over the repairs on the same day, re-fetched all three sources, and found every repair supported; six phrases were tightened to the sources' own wording: the Lazard solar-plus-storage figure is now given as Utility Dive's citation of a national April 2023 estimate, unsubsidized, rather than as a feature of UAMPS's market; the CATF quotation on utilities' hesitation now carries its full clause ("that had not even initiated the regulatory process"); CATF's first-customer list is given as "may be better suited" rather than "best placed"; CATF's count is "more than 50 SMRs", not "SMR designs"; E&E News's "several" members withdrew, not all; and the purpose CATF gives for large order books ("to support the scale-up of the nuclear market") replaces a fleet-carries-the-cost mechanism no source states.