InvestmentsNuScale Power and Utah Associated Municipal Power Systems (UAMPS) cancelled the Carbon Free Power Project (CFPP) in November 2023. It would have been the first commercial SMR plant in the United States. ⚡ The plan was six 77 MWe NuScale Power Modules, 462 MWe in total, on the Idaho National Laboratory site near Idaho Falls, with E&E News reporting a 2029 start date. Nearly three years later, the project is still the best case study in the sector. Most investors remember “costs rose” and stop reading. The useful parts sit in contract clauses, subscription math and a regulatory approval that did not produce a buyer.
How the Idaho project came apart
The project shrank before it died. 📉 UAMPS first planned 12 modules and 720 MW, then cut the plant to six. The target price for power moved from $58/MWh in 2020 to $89/MWh in January 2023, a rise of 53%. NuScale’s own cost update pointed to steel and electrical equipment, noting that carbon steel piping and plate prices were up more than 50% since 2020. Inflation did real damage here, but it was not the whole story.
The numbers below are the ones worth keeping on a notecard. 🧾
Reactor: six NuScale modules at 77 MWe each, 462 MWe total
Owner and site: CFPP LLC, a UAMPS subsidiary, at Idaho National Laboratory
Price target: $58/MWh in 2020, $89/MWh by January 2023
Total cost: $9.3 billion including financing, with $5.1 billion net to UAMPS after DOE and other support
DOE backing: a cost-share award of up to $1.4 billion, with $232 million paid out by cancellation
Here is the part that stings. A project can survive a cost overrun. It cannot survive a customer who has a legal right to leave. 🚪 Ask yourself how many SMR announcements you have read this year that mention an exit clause at all.
Lesson one: the price target is a trigger, not a talking point
In December 2022, NuScale and Fluor delivered an economic competitiveness test to CFPP LLC. The calculated cost of power exceeded the $58/MWh target. Under the amended agreement, that failure gave UAMPS members the right to leave, and they had until March 1, 2023 to decide. Enough of them stayed. NuScale’s Form 10-Q also says it could have owed CFPP LLC money had the project ended at that point. 💸
The reset to $89/MWh came with a second tripwire. If the test result exceeded $89, or if subscription fell below 370 MWe, members could exit again. That is a customer-held option, and options get exercised. I think this is the least appreciated fact about the whole saga. 🔍
Before you trust any SMR contract summary, check these:
Who sets and measures the price ceiling
Who can walk away, and until what date
What the customer gets reimbursed on exit
Whether the vendor owes anything when the test fails
Lesson two: subscriptions are the project
UAMPS is a nonprofit utility of roughly 50 municipal members, and each member chose its own level of participation. CFPP targeted 80% subscription by the end of 2023. NuScale CEO John Hopkins told the ANS Winter Conference that that threshold looked out of reach despite “significant efforts” by both parties. 📊 Municipal buyers also carry a mandate to buy the cheapest power they can. Atomic Insights argued that access to local coal and tax credits that pay out only after completion made nuclear a harder sell than wind. 🌬️
The new deals move that risk to bigger balance sheets. Compare the three structures:
CFPP: municipal members subscribe, an 80% target, and the buyers carry the cost
TVA: ENTRA1 Energy would finance, own and operate six plants and sell power to the Tennessee Valley Authority, up to 6 GW or 72 modules, still short of a definitive power purchase agreement
Romania: RoPower Nuclear plans six 77 MWe modules at Doicesti, an estimated $6-7 billion, with the state funding only the first unit
That is progress. 🏗️ It is also not the same as a signed contract. Which of those three structures would you rather hold the risk on? Decide before you buy the stock.
Lesson three: regulatory approval is not a customer
The NRC certified NuScale’s 50 MWe design in July 2022. CFPP died sixteen months later anyway. In May 2025, the NRC approved the uprated 77 MWe design, which DOE says it has backed with more than $575 million. 🏛️ NuScale still describes itself as the only SMR developer with NRC design approval. That is a real moat. It is not revenue.
The money tells the same story. NuScale ended the second quarter of 2026 with $1.9 billion in cash and investments, but reported quarterly revenue of only about $75,000, down roughly 99% from a year earlier according to Earnings Whispers. 💰 The company’s results mention no definitive power purchase agreement signed that quarter. If you’re tracking the nuclear market professionally, SMRbrief Pro gives you the structured database to go deeper than any single article can. 📚
A rough scorecard as of August 2026, so check for newer filings:
Holds: two NRC design approvals, $1.9 billion in liquidity, a Romanian final investment decision
Lacks: a binding TVA purchase agreement, a delivered module, a plant under construction in the United States
Mixed: the Romania decision carries conditions attached to the Nuclearelectrica shareholder vote
What to watch before the next headline
Skeptics have a point, and so do the optimists. 🎯 The Romania vote on the final investment decision is the nearest thing NuScale has to a firm order. The Power Magazine analysis flags the caveats, and the Prime Minister cautioned that it does not mean immediate groundbreaking. The TVA program is bigger, but ENTRA1 and TVA were still negotiating as of NuScale’s August 2026 results. 🔭
Watch these, in order:
A signed, definitive ENTRA1 and TVA power purchase agreement
Satisfaction of the conditions on the RoPower final investment decision
Manufacturing readiness for the first module, including the safety-system contract NuScale awarded to Paragon
The pace of cash burn against that $1.9 billion
Which would you accept as proof that Idaho was a one-off: a signed TVA contract, a poured foundation in Romania, or a finished module leaving a factory?