Why SMR PPAs require a new approach

Bridging this gap will be essential to securing energy supply for the fast-evolving needs of digital infrastructure.

This article is a collaborative effort by Peter Hans Hirschboeck, Lisa Nguyen, and Mark Cira, representing views from impactECI

---

U.S. data center power demand is projected to triple by 2030, outpacing grid capacity and straining existing clean energy options. While SMRs won’t be commercially deployed in significant numbers until the 2030s, they represent a critical long-term solution for meeting rising digital power demands. Now is the moment to act. Developers, utilities, and energy buyers must come together to create new commercial structures, specifically, PPA frameworks that reflect the unique capital cycles, risk profiles, and deployment timelines of advanced nuclear. Bridging this gap will be essential to securing energy supply for the fast-evolving needs of digital infrastructure.

Surging demand-widening supply gap

U.S. data center electricity demand is entering a period of exponential growth, expected to triple between 2023 and 2030¹. Driven by AI adoption, digitalization, and cloud services proliferation, data center consumption could rise to 580–800 TWh²/year at CAGR 15-21%, potentially accounting for up to 12% of total U.S. electricity use in high-demand scenarios. As a result, the grid could face a 300–600 TWh increase in load by 2030, which translates to 65–90 GW at full capacity or 95–130 GW at a 70% capacity factor. The resulting load profile of these AI data centers rivals that of the nation’s largest industrial sectors.

Over the same period, total U.S. electricity generation is projected to increase by only ~ 300 TWh, from 4,179 TWh in 2023 to 4,480 TWh by 2030, a 1.4% CAGR. In the low-case scenario, this increase would barely meet the incremental increase from data centers. In the high case scenarios, it would fall short- even before accounting for electrification across other sectors, such as transportation and buildings. While energy efficiency gains in those sectors may offset some demand, the underlying imbalance remains a structural concern.

Unlike typical commercial facilities, data centers operate at some of the highest load factors among all customers, meaning their energy usage remains consistently elevated around the clock. As a result, meeting this demand will require not only significant grid infrastructure upgrades, but also the deployment of firm, dispatchable generation to supplement intermittent renewables and ensure round-the-clock reliability.

This broadening shortfall is being compounded by systemic issues to grid expansion. Lengthy and complex permitting processes, slow regulatory approvals, transmission easements, and construction delays, while new loads, such as data centers and advanced manufacturing, are connected faster than new generation, further delaying project timelines. Meanwhile, supply chain hurdles across the generation and transmission (G&T) sector are growing more acute. Critical components such as generators, turbines, conductors, transformers, and switchgear remain in short supply, with lead times extending into multiple years. Labor shortages further exacerbate the issue. Moreover, limited land availability in high-density regions, such as Northern Virginia and New York, undermines the viability of onsite renewable deployment. Together, these challenges signal a looming power crunch, one that could delay the buildout of digital infrastructure.

Small modular reactors (SMRs) offer a dispatchable, carbon-free option

SMRs offer a compelling path forward to meet surging digital power needs with carbon-free, reliable electricity. SMRs promise high-capacity factors, delivering³, load-following power that complements the variability of renewables- providing stability when wind and solar output falls short. When configured with N+1 or N+2 redundancy, integrated battery energy storage systems (BESS), and/or robust backup generation, SMRs can further enhance system availability and uptime, ensuring uninterrupted performance⁴. In addition, SMRs can be deployed behind the meter or in microgrid configurations, physically and electrically isolating data centers from the broader transmission grid. This not only mitigates potential cross-subsidization, where infrastructure upgrades that otherwise would have been needed are avoided thanks to on-site generation, but also addresses power quality concerns by containing voltage fluctuations, harmonics, and other disturbances at the source. Co-locating data centers with existing or swiftly built new nuclear power plants enables synergies in security capex and opex, optimizing costs and delays by eliminating the need for new transmission infrastructure and grid interconnection studies.

Intermittent renewables fall short of always-on digital power needs

While solar and wind are cost-competitive and carbon-free, they are inherently non-dispatchable, generating electricity only when weather conditions allow. Typical capacity factors are ~29% for solar and up to ~60% for wind across a well-diversified portfolio⁵, which cannot meet the continuous, high-reliability power requirements of mission-critical loads such as hyperscale data centers. Delivering solely through battery storage integration (BESS) would require an uneconomical scale of storage investment. Meanwhile, reliability challenges compound this issue. In the SPP peak ELCC report, utilities can count on only 15-25% of wind capacity and 36-62% of solar capacity to meet peak demand⁶. Regardless of grid and land constraints, renewables alone cannot provide the firm power required by data centers. Without a scalable source of dispatchable, clean electricity, the supply-demand imbalance will only widen.

Policy pivot: declining renewable subsidies, rising nuclear support under the One Beautiful Bill Act (OBBB)

Recent legislative shifts under the OBBB, signed on July 4th, 2025, pose substantial challenges to renewable energy growth by curtailing key IRA tax incentives, including:

Accelerated credit phase-out: ITC and PTC for wind and solar projects now expire for facilities placed into service after 12/31/2027, with only a one-year window from enactment to begin construction.

Foreign-entity restrictions- new limitations prohibit projects involving designated foreign entities from accessing remaining credits, narrowing eligibility and complicating strategic partnerships. These shifts significantly reduce renewable deployments, with Princeton analysis projecting a capacity reduction of 160 GW for wind and 140 GW for solar by 2035.

Meanwhile, the OBBB also offers a notable tailwind for nuclear energy and for SMRs. The legislation provides targeted carve-outs allowing both new and existing nuclear facilities, including SMRs, to retain eligibility for clean electricity production and investment tax credits long after renewable incentives have phased out. These provisions are poised to direct billions in subsidies toward nuclear projects, significantly accelerating the deployment and financial viability of advanced reactors even as renewable incentives diminish.

The commercialization of SMRs is gaining momentum- yet remains stalled by industry constraints

Globally, the SMR landscape is diverse, with dozens of designs under development across various reactor types⁷. Developers such as GE Hitachi (BWRX-300), X-energy (Xe-100), TerraPower (Natrium), and Holtec (SMR-160) are advancing their designs, targeting FOAK deployments in the late 2020s to early 2030s. Meanwhile, China built and connected the “Linglong One” (ACP-100) to the grid, the world’s first onshore commercial SMR⁸. Yet, U.S. SMR commercialization remained stalled. Despite growing policy support, progress is hindered by four structural barriers: licensing delays, persistent cost escalations, financing shortfalls, and fuel supply risks.

Licensing remains slow: As of 2025, no modern SMR has reached full operational maturity in the U.S. Further, the U.S Nuclear Regulatory Commission (NRC) has limited experience in licensing non light-water reactor designs. SMR developers face a strategic choice between two distinct licensing pathways that fundamentally shape project timelines, risks, and flexibility. The 10CFR52 framework requires standardized design certification through a combined license (COL), providing regulatory certainty and long-term scalability but entailing lengthy, multi-year reviews. NuScale’s VOYGR SMR exemplifies this path, being the first and only design to achieve U.S. certification after six-year⁹ certification process, finalized in 2025 with its uprated 77 MWe module. Conversely, many advanced reactor developers, such as X-Energy, pursue the 10CFR50 two-step process, which bypasses design certification, enabling faster construction permit approvals, often within 12 to 24 months¹⁰and greater design flexibility during construction. While this accelerates FOAK deployments, it introduces regulatory uncertainty, as operating licenses remain required and the pathway has seen limited recent use for new designs. Ultimately, SMR developers must balance the long-term benefits of standardized certification and regulatory predictability under 10CFR52 against the shorter timelines and adaptable design approach afforded by 10CFR50, with each choice reflecting differing risk-return profiles for market entry and scale.

Historical nuclear cost estimate increases have spooked investors: For example, NuScale’s first project, the Carbon-Free Power Project in Idaho, was cancelled, and X-energy’s project estimates rose from $2.5b to $5.75b¹¹. Labor and material cost inflation, such as a 42% surge in specialty steel prices¹², adds further uncertainty to long-term budgets. However, much of the gap stems from early-stage estimates that lacked detailed engineering and applied undisciplined contingency allowances. As designs move toward licensing, more accurate cost baselines are emerging but those are often compared to earlier un-escalated figures, creating the appearance of steep increases. The significant lack of experience with advanced nuclear construction in the US exacerbates these forecasting challenges, reinforcing investor skepticism.

Financing structures remain misaligned with SMR risk profiles: The conventional delegation of risk in energy infrastructure projects between developers, utilities, EPCs, and OEMs does not adequately address the unique challenges of SMRs. As a result, many projects struggle to advance beyond the planning phase.

Distinct capital risk profile
For developers and utilities, the primary risks stem from capital intensity, multi-year permitting cycles, and delivery uncertainty. SMR projects typically involve 8–10 years of design, construction, and commissioning, followed by 40+ year asset lives. This long capital lock-up requires pricing models that account for inflation volatility, interest rate variability, and multi-decade discounting assumptions. To manage these risks, utilities often seek 25–30-year offtake agreements to amortize these risks, tenors that rarely align with the shorter planning horizons of commercial offtakers, especially in fast-moving sectors like data and AI infrastructure.

Buyer – project developer mismatch
Hyperscalers, colocation providers, and key potential offtakers prioritize flexibility, scalability, and responsiveness to shifting demand driven by AI workloads and digital infrastructure growth. Their lease and investment cycles typically span 10–15 years, far shorter than traditional nuclear PPA terms. Location mismatches compound the issue. If the SMR is not co-located with the data center, buyers face basis risk, transmission congestion, and the potential for costly and time-consuming grid upgrades. For data center operators investing billions into new campuses, energy delivery delays can tie up capital and derail operational timelines. Long-term fixed-quantity PPAs are difficult to underwrite without volume flexibility, milestone-based pricing, or delivery guarantees.

Risk allocation gap among utility, EPC, and OEM stakeholders

Beyond buyer-developer friction, misaligned risk allocation among utilities, EPCs, and OEMs remains a critical hurdle. Utilities, in vertically integrated markets, seek guaranteed delivery timelines and cost certainty; EPCs face limited ability to absorb multi-year construction risk; and OEMs, often early-stage SMR vendors, lack the balance sheets to backstop large-scale guarantees. Without clear frameworks for risk sharing and project-level insurance mechanisms, achieving financial close remains difficult, particularly in FOAK deployments.

Fuel supply is another primary barrier: Most non-light water reactors rely on high-assay low-enriched uranium (HALEU), yet current U.S. domestic output stands roughly at 0.9 tons p.a- far below the~ 50 tons p.a of HALEU¹³ projected to be needed by 2035. The U.S. sourced ~ 27% of its enriched uranium from Russia, which controls 44% of global supply¹⁴ - introducing geopolitical exposure. Enrichment caps under the Prohibiting Russian Uranium Imports Act could reduce efficiency by up to 18% for some SMRs. While the 2027 waiver provides short-term relief, long-term fuel uncertainty continues to stall investment, disrupt offtake discussions, and elevate risk premiums. To address this, developers have pursued different strategic sourcing strategies to mitigate HALEU risks. Terra Power, for example, has partnered with ASP Isotopes to develop an enrichment facility in South Africa, targeting 15 tons p.a by 2027 and scaling to 150 tons from 2028. This complements TerraPower’s broader fuel strategy, anchored by domestic initiatives and partnerships with Centrus and Framatome, to build a more resilient, globally distributed supply chain

Executive action and DOE support address key barriers—but gaps remain

Since 2020, the DOE’s Advanced Reactor Demonstration Program (ARDP) co-funds two flagship projects targeting deployment by 2030: Natrium/ GE-Hitachi, a 345 MWe located at a retiring coal site in Wyoming and Xe-100/X-energy, a 320 MWe sited at a Dow facility on the Gulf Coast. Each receives up to $1.9B in federal funding, matched by private capital. Both projects advance SMR commercialization by scaling HALEU fuel infrastructure, demonstrating regulatory pathways (via NRC Part 50). ARDP also supports five additional designs and three early-stage concepts, ensuring a diverse, future-ready reactor pipeline.

The federal government has committed an additional $900m, including expanded access to the Department of Energy’s Loan Programs Office (LPO), specifically targeting Generation III+ light-water SMRs for near-term commercialization. These funds are intended to de-risk early-stage development and support FOAK projects, addressing financing shortfalls that have historically limited progress. In parallel, the May 2025 Executive Orders mark a notable shift in federal posture, targeting two additional structural challenges:

Regulatory acceleration: To streamline long and uncertain approval timelines, the NRC is directed to expedite licensing for advanced reactors, particularly those previously tested at DOE facilities and to issue final construction and operating license decisions within 18 months.

Fuel supply security: To mitigate fuel-related constraints, the DOE is instructed to release 20 metric tons of HALEU from federal reserves and expand domestic enrichment, conversion, and reprocessing capacity.

These actions represent important steps toward reducing structural risk across licensing, financing, and fuel availability. However, they stop short of fully closing the commercial gap. Challenges around project cost certainty, capital intensity, and long-term offtake misalignment remain unresolved. Unlocking large-scale SMR deployment will require deeper capital stack innovation, standardized risk-sharing frameworks, and flexible contracting models tailored to the unique economics of SMRs.

Emerging SMR PPA models: rebalancing risk, capital, and flexibility

Unlike standardized bilateral PPAs used in wind and solar, SMR projects involve a more fragmented ecosystem, spanning developers, utilities, corporate offtakers, regulators, fuel suppliers, and financiers. The cancellation of NuScale’s Carbon-Free Power Project¹⁵ highlighted the economic and delivery risks of (FOAK) deployments, reinforcing the need for tailored risk-sharing mechanisms across all parties.

Addressing these challenges requires a rethinking of commercial structures, anchored in contractual innovation that balances risk, flexibility, and bankability. Leading developers and buyers are exploring a suite of options, including milestone-based payments, indexed pricing, performance guarantees, virtual PPAs, equity participation, and government-backed risk buffers. A few example mechanisms include:

Risk sharing frameworks: Traditional PPAs often place a disproportionate burden on developers for delays, cost overruns, or regulatory uncertainty. SMR contracts are increasingly embedding shared responsibility models, where buyers contribute during pre-construction or assume partial risk tied to macroeconomic events. This may include indexed price escalation, capital cost bands, or milestone-triggered cost adjustments to reflect changing risk as projects progress.

Delivery contingency clauses: Given SMRs’ multi-year build cycles and regulatory complexity, offtakers require protection against uncertain start dates. Leading contracts now include “excusable delay” provisions linked to licensing milestones, force majeure tied to NRC timelines, and liquidated damages capped by delivery windows. These mechanisms provide buyers and developers clarity on the downside while maintaining alignment with development realities.

Multi-unit optionality for scalability: To bridge the mismatch between asset lives, financing tenors, and 10–15-year buyer planning horizons, contracts are adopting modular commitment models. Buyers may start with 1–2 units under a defined tranche, with options to expand contingent on performance, demand growth, or permitting progress. This phased approach provides flexibility while enabling developers to build scale.

Case examples: innovation in action

Amazon and X-energy: aligning capital and flexibility through phased deployment¹⁶
Amazon’s $500m partnership¹⁷ with X-energy aims to deliver up to 960 MW of SMR capacity, beginning with four 80 MW Xe-100 units. The agreement includes both reactor and fuel investments, supporting the TRISO-X HALEU facility to de-risk supply. The model blends scalability with contractual flexibility, aligning generation delivery with evolving data center demand.

Google and Kairos: milestone-based contracting to reduce exposure¹⁸
In contrast, Google’s agreement with Kairos Power is built around a Master Plant Development Agreement targeting 500 MW by 2035. Rather than committing to fixed delivery, the contract ties deployment to clearly defined regulatory and technical milestones. The PPA includes ancillary services and carbon-free attributes, supporting Google’s 24/7 clean energy goals, while decoupling development risk through a structure where Kairos retains operational control.

Dominion and GE Hitachi: cost efficiency through design repeatability¹⁹
Dominion Energy’s collaboration with GE Hitachi focuses on the BWRX-300 platform, leveraging standardized design and industrial-grade modular construction to reduce capital costs by up to 60% versus traditional nuclear builds. The approach builds on the licensing foundation of GE’s ESBWR design and benefits from shared governance with Ontario Power Generation and TVA.

Summary

Digital power demand is rising faster than the grid can respond. Renewable deployment alone won’t close the gap, especially under today’s siting, permitting, and transmission constraints. SMRs can help fill that void. But commercialization won’t succeed without a new playbook. Success will require new commercial models that address fuel readiness, delivery timelines, supply chain risks, financing challenges, and multi-party risk. Early partnerships are setting precedent, but broader adoption will require coordinated action across the SMR value chain.

At impactECI, we help clients translate these complexities into action, designing fit-for-purpose PPAs, building shared-risk frameworks, and aligning energy infrastructure with digital-era power needs. Whether you're an SMR developer, buyer, or investor, now is the time to structure for scale.

For more information, reach out to hello@impactECI.com

The views and opinions in these articles are solely of the authors. They are offered to stimulate thought and discussion and not as legal, financial, accounting, tax or other professional advice or counsel.

1. https://eta-publications.lbl.gov/sites/default/files/2024-12/lbnl-2024-united-states-data-center-energy-usage-report_1.pdf

2. https://eta-publications.lbl.gov/sites/default/files/2024-12/lbnl-2024-united-states-data-center-energy-usage-report_1.pdf

3. https://css.umich.edu/publications/factsheets/energy/nuclear-energy-factsheet

4. https://www.sciencedirect.com/science/article/pii/S2590174523000533

5. https://www.spglobal.com/market-intelligence/en/news-insights/research/2022-monthly-us-solar-capacity-factors-underscore-winter-doldrums

6. https://www.spp.org/documents/72346/2024%20spp%20elcc%20wind%20solar%20&%20esr%20report.pdf

7. https://aris.iaea.org/publications/SMR_catalogue_2024.pdf

8. https://www.sustainability-times.com/energy/china-leaves-west-in-the-dust-its-small-nuclear-reactor-leap-puts-beijing-years-ahead-in-the-future-of-clean-energy/

9. NRC.gov

10. https://www.businesswire.com/news/home/20250616282344/en/NRC-Publishes-18-Month-Review-Schedule-for-Xe-100-Construction-Permit-Application

11. https://ieefa.org/s/SMR-too-risky-too-slow-too-pricy

12. procurementresource.com

13. https://www.energy.gov/ne/articles/what-high-assay-low-enriched-uranium-haleu

14. https://www.spglobal.com/commodity-insights/en/news-research/latest-news/electric-power/111524-russia-temporarily-bans-exports-to-us-of-enriched-uranium-after-us-curbed-imports

15. www.nytimes.com

16. https://x-energy.com/media/news-releases/amazon-invests-in-x-energy-to-support-advanced-small-modular-nuclear-reactors-and-expand-carbon-free-power

17. https://finance.yahoo.com/news/amazon-invests-x-energy-support-130000469.html

18. https://neutronbytes.com/2024/10/15/google-inks-power-purchase-deal-for-seven-kairos-reactors/

19. https://www.ge.com/news/press-releases/ge-hitachi-announces-dominion-energy-investor-bwrx-300-small-modular-reactor