The economics of small modular reactors remain the single biggest question in the nuclear renaissance. SMR proponents promise lower capital costs through factory fabrication, shorter construction timelines, and learning-curve effects from serial production. Critics point to NuScale's CFPP cancellation — where costs escalated from $5.3B to $9.2B — as evidence that small reactors face the same cost diseases as large ones. The truth is nuanced: first-of-a-kind SMRs will likely cost $80-150/MWh, competitive with large nuclear but not with solar or wind. The cost advantage emerges only with Nth-of-a-kind production, where estimates drop to $50-80/MWh. This page breaks down the real numbers.
| POWER SOURCE | LCOE ($/MWH) | CAPACITY FACTOR | DISPATCHABLE | CO2 (G/KWH) | NOTES |
|---|---|---|---|---|---|
| Utility-Scale Solar (PV) | $30-50 | 20-30% | No | 0 | Cheapest source; intermittent, needs storage |
| Onshore Wind | $25-50 | 25-45% | No | 0 | Low cost; intermittent, location-dependent |
| Natural Gas (CCGT) | $40-75 | 85-90% | Yes | 350-400 | Low capex, high fuel/carbon exposure |
| Large Nuclear (new-build) | $80-180 | 90-93% | Yes | 0 | Vogtle 3&4: ~$140-180/MWh realized |
| SMR (FOAK estimate) | $80-150 | 90-95% | Yes | 0 | First-of-a-kind; high uncertainty |
| SMR (NOAK target) | $50-80 | 90-95% | Yes | 0 | Requires serial production & learning |
| Solar + 4h Battery | $55-85 | ~70% | Partial | 0 | Higher capacity factor but not 24/7 |
| Offshore Wind | $60-100 | 40-55% | No | 0 | Higher output but expensive installation |
| REACTOR | DEVELOPER | COST/KW | TOTAL FOAK | TECHNOLOGY | NOTES |
|---|---|---|---|---|---|
| BWRX-300 | GE Hitachi | $3,000-5,000 | N/A (Darlington est. $3-5B CAD) | Light water BWR | Simplified design, lowest capex target |
| Natrium | TerraPower | ~$11,600 (FOAK) | $4B (Kemmerer) | Sodium fast reactor | Includes DOE $2B cost-share |
| Xe-100 (4-pack) | X-energy | $5,000-8,000 (est.) | $2-3B (est. 320 MWe) | HTGR | TRISO fuel adds cost but enhances safety |
| AP1000 | Westinghouse | ~$13,000 (realized) | $35B (Vogtle 3&4, 2.2 GW) | Large PWR | US FOAK; Korea builds at ~$5,000/kW |
| NuScale VOYGR (cancelled) | NuScale | ~$20,000 (final est.) | $9.2B (462 MWe CFPP) | Light water PWR | Cost escalated 74% before cancellation |
The NuScale Carbon Free Power Project (CFPP) is the most important case study in SMR economics. It demonstrates that regulatory-first does not guarantee commercial success, and that small reactors are not immune to the cost escalation that has plagued large nuclear projects. Read the full NuScale story →
First and only SMR to receive NRC design certification
Carbon Free Power Project with UAMPS in Idaho. Target: 462 MWe (6 x 77 MWe). Initial cost: $5.3B
First cost increase. Price per MWh rises to $89/MWh
DOE conditional commitment of $1.4B. Price hits $89/MWh then $100+/MWh. Municipal subscribers begin withdrawing.
UAMPS terminates agreement. Not enough subscribers at escalated price. NuScale writes down $471M. Stock drops 80% from 2022 peak.
Romania RoPower (6 modules, 462 MWe), Poland deployment. NRC SDA for uprated 77 MWe design approved May 2025.
Trading on NYSE (SMR). Market cap ~$3B. No US deployment. International pipeline: Romania, Poland, Czech Republic. Cautionary tale for SMR economics.
Nuclear LCOE is dominated by upfront construction costs. Unlike gas plants (where fuel is 60-70% of LCOE), nuclear fuel is cheap ($5-7/MWh) but the plant is expensive to build.
Long construction timelines (5-10 years) accumulate billions in IDC. A project at 8% cost of capital adds ~50% to overnight cost over 7 years.
NRC licensing, safety analysis, environmental review, and quality assurance add years and billions. US regulatory costs are among the highest globally.
FOAK plants face design finalization during construction, supply chain establishment, and workforce training — all of which inflate costs vs standardized designs.
Nuclear construction requires specialized welders, electricians, and engineers. The US nuclear construction workforce has atrophied since the 1980s build-out.
Nuclear-grade components (reactor vessels, steam generators, fuel assemblies) have limited suppliers, creating bottlenecks and pricing power.
SMRs are not cheap today — and anyone claiming otherwise is selling something. First-of-a-kind SMR power will likely cost $80-150/MWh, more expensive than solar ($30-50/MWh), wind ($25-50/MWh), and natural gas ($40-75/MWh). The value proposition is not cost competitiveness on day one — it is the ability to deliver 24/7 carbon-free dispatchable power at a 90%+ capacity factor with zero emissions, something no other source can match. The real question is whether serial production drives NOAK costs down to the $50-80/MWh range. NuScale's failure shows the risk of cost escalation is real. But the hyperscaler nuclear deals — Meta's 6.6 GW, Microsoft's TMI restart, Amazon's Xe-100 fleet — suggest the market believes the economics will work at scale. For investors and policymakers, the lesson is clear: bet on the technology pathway with the clearest route to Nth-of-a-kind production, not the one with the lowest FOAK estimate.