Nuclear energy financing must undergo a parallel shift alongside improvements in technology in order to witness the second nuclear renaissance.
When people talk about America’s “second nuclear renaissance,” they usually focus on technology. They point to the small modular reactors (SMRs) and microreactors (MRs) now being authorized by a number of government agencies, the rapid spread of data centers to fuel artificial intelligence, the urgency of decarbonizing industry through technological innovation, and the geopolitical upheavals that have exposed how fragile global energy technology supply chains can be.
Those developments certainly matter. But in a soon-to-be-published article in the George Washington International Law Review, David Beckstead and I found that the future of nuclear power in the United States will depend as much on the emergence of a new and viable financing architecture for nuclear power as much as on new reactor designs. The most consequential innovations in nuclear energy right now are not only occurring in laboratories and engineering firms but also in loan agreements, power-purchase contracts, and balance sheets.
What we found and detail in the article is that the United States has quietly entered a new era in which nuclear projects are increasingly built around early commitments from large private offtakers — i.e., data center operators, industrial conglomerates, and government agencies — rather than around utility planning or state-mandated procurement. This structure, which we describe as Finance-to-Offtake (FTO), represents a fundamental inversion of the historical model of nuclear development. And it is this inversion, rather than any particular reactor technology, that may ultimately determine the pace and scale of America’s nuclear energy revival.
A Shift in Nuclear Development Financing
For most of the last half-century, nuclear development followed a predictable script. A vertically integrated utility would propose a reactor, a state public utility commission would approve the recovery of construction costs through electricity rates, and investors would finance construction with the confidence that cost overruns, however inevitable, would eventually be absorbed by ratepayers. This model worked in an era of regulated monopolies and relatively stable electricity demand. It made nuclear power feasible by socializing financial risk across millions of captive customers.
But in wide swaths of the country, electricity is now delivered through competitive wholesale markets in which generators live or die on hourly energy prices. Nuclear plants cannot reliably be built in the hope that future electricity prices will repay a multi-billion-dollar investment. Even in states that retain regulated utilities, the political tolerance for massive rate increases has eroded. The experience of the Vogtle expansion in Georgia, delivered years late and at a cost exceeding $30 billion, has only reinforced the sense that traditional nuclear financing is untenable.
In this environment, the question is not simply who makes the reactors or how safe and modular they are. It is who will buy the power and on what terms.
It is here that the United States is witnessing its most important shift. As David and I found, the rise of hyperscale computing has transformed the shape of electricity demand.
Artificial-intelligence data centers now require hundreds of megawatts of continuous power per site, and in some cases multiple gigawatts when campuses are aggregated. Their demand does not fluctuate with weather or gas prices. They cannot rely on intermittent renewables. And they cannot wait seven years for a place in a transmission-interconnection queue. For these firms, the old utility-centric market design plays almost no role in their planning. What they need is reliable
power, delivered directly, on predictable terms, as quickly as possible.
Nuclear Reactors and Technology Companies
This demand profile has made large technology companies a new kind of anchor customer — one with both the creditworthiness and the appetite to commit to long-term, fixed-price power from nuclear facilities. The same is true for industrial firms, particularly those involved with chemicals, steel, and materials processing, which need continuous high-temperature heat and electricity that wind and solar simply cannot provide. Even federal agencies, especially the Department of Defense, have become potential anchor offtakers for microreactors designed for energy resilience at remote facilities.
These customers are not merely purchasers of power at the end of the development cycle. They are beginning to shape the financing itself. In three recently announced transactions, Meta, Amazon, and Microsoft each signed long-term power-purchase agreements with existing nuclear reactors — agreements large enough to influence critical decisions about reactor operation and, in Microsoft’s case, even the restarting of a previously closed reactor at Three Mile Island Unit 1. These were not symbolic gestures. They mark the first time in US history that commercial hyperscalers have contracted directly for nuclear power at this scale.
But the more interesting shift is happening not with existing reactors but with projects that are still conceptual or in early development. Amazon has entered into a joint development agreement with X-Energy that is tied to the eventual deployment of Xe-100 high-temperature gas reactors in Washington state, an arrangement centered on Amazon’s future offtake of the power. Dow Chemical is partnering with X-Energy on a four-unit SMR project in Seadrift, Texas — one of the first nuclear projects in US history driven almost entirely by the power needs of a single industrial facility. The US Air Force has awarded long-term offtake contracts for the Oklo Aurora microreactor to power Eielson Air Force Base in Alaska. And Google is exploring similar arrangements with Kairos Power’s next-generation reactors.
In every one of these cases, the logic is consistent: the customer comes first, and the customer’s commitment becomes the basis for financing the reactor. This is the inversion that defines the Finance-to-Offtake Model. The offtake agreement is no longer a late-stage component of a utility’s procurement plan; it is the linchpin around which the project exists.
Implications of the Shift in Financing Methods
The implications of this shift cannot be overstated. In traditional nuclear finance, developers had to attract billions of dollars of capital before they could credibly approach customers. Today, the presence of a creditworthy customer is what unlocks capital in the first place. Investors are willing to enter nuclear projects only when a high-quality offtaker has already committed to purchase power over a multi-decade horizon at a stable price. The offtaker’s balance sheet becomes the de-risking mechanism that replaces the old state-regulated rate base.
This pattern, where early customer commitment drives project financing, has deep analogues in the recent development of other energy industries. In the early days of the US shale gas boom, LNG export terminals were financed only after long-term offtake agreements with foreign buyers. Renewable projects, such as wind and solar, became bankable once corporate power purchase agreements (PPAs) became a standardized instrument for managing market risk. Even major pipelines and petrochemical complexes rely on early anchor-tenant commitments before any construction capital is deployed. Nuclear is now beginning to migrate toward this same demand-first logic.
Finance-to-Offtake also reflects another important reality: the grid itself has become a bottleneck. Interconnection queues in a number of US regional transmission organizations (e.g., Pennsylvania-New Jersey-Maryland (PJM) and Midcontinent Independent System Operator (MISO)) are so congested that even straightforward renewable projects can face delays of five to seven years before receiving permission to connect. For datacenter developers and industrial firms, behind-the-meter nuclear solutions — i.e. reactors that are purpose-built to power their facilities — are often more attractive than waiting for grid infrastructure that may never arrive. Many of the new SMR and MR designs are physically small enough to be collocated with industrial complexes or data campuses, allowing direct, committed supply arrangements that bypass the traditional grid entirely.
This, too, shifts the center of gravity of nuclear finance away from utilities and toward large end-users. The new nuclear market is not about adding generic capacity to a regional grid. It is about building firm power or process heat systems for specific customers with specific needs. That specificity makes early offtake not only possible but also essential.
Some observers argue that these developments will eventually pave the way toward true project-financed nuclear projects, in which stand-alone special-purpose entities borrow against the cash flows of the project itself, without recourse to the balance sheets of the developer or utility.
Is This New Model of Nuclear Energy Financing a Possibility or Just a Dream?
David and I think that is plausible, and we note that this model has been used successfully for multiple nuclear projects in Finland, but in the United States we think that this is likely only after several further stages of maturation. Early SMR and MR deployments are still too risky, too bespoke, and too first-of-a-kind to attract non-recourse debt at scale. For now, they remain financed through combinations of developer equity, federal cost-share, loan guarantees, and vendor participation. But as deployments become repeatable, and as early reactors demonstrate predictable costs and performance, the FTO model may be the mechanism through which nuclear finance begins to resemble the financing structures that propelled the growth of wind and solar in the 2000–2020 period.
If that happens, we believe the first true project-financed nuclear reactors in the United States will not be built by regulated utilities at all. They will likely be built in places like Texas, where private bilateral contracting is common and where industrial and digital loads are expanding at a pace unmatched anywhere in the country. They may also emerge in the Pacific Northwest, where public power entities like Energy Northwest can aggregate offtake commitments and issue revenue-backed bonds. And they may appear in states with aggressive clean-energy procurement mandates, where long-term contracts for “24/seven clean power” could serve the same role that Contracts-for-Difference played in making the UK’s Hinkley Point C project possible.
But regardless of where they appear, the decisive element will be the same. The market conditions that once made nuclear power plants almost impossible to finance are shifting because large private actors are stepping into the role utilities once played. They are offering the long-term revenue certainty that nuclear projects require, not because they seek to serve a public grid, but because their own operations demand levels of electricity that only firm, high-availability, zero-carbon sources can provide.
This is why the FTO trend is so important. It is not simply a novel contractual arrangement; it is a structural reorientation of the nuclear industry itself. It aligns nuclear deployment with the realities of modern electricity demand and modern capital markets. It transforms nuclear energy from a public utility asset into an industrial or digital infrastructure asset. And it provides a path for nuclear energy to expand even in regions where utilities, regulators, and electricity markets are not naturally inclined to support it.
If the second nuclear renaissance does take hold in the United States, it will likely not be because any particular SMR design became the winning technology or because political leaders suddenly deduced that nuclear power should be subsidized. It will be because large buyers of electricity — i.e., the companies and institutions that need it most — discovered that they could actualize the delivery of clean, reliable, committed power by promising to buy it first.
That is the logic at the heart of the Finance-to-Offtake Model. And if early trends are any guide, it may become the defining logic of nuclear industry development in the United States in the decades ahead.
About the Author: Daniel Joyner
Daniel Joyner is the Elton B. Stephens Professor of Law at the University of Alabama School of Law and the founder and principal at Prometheus Nuclear LLC. Prior to joining the Alabama Law faculty in 2007, Dr. Joyner taught for four years on the faculty of the University of Warwick School of Law. He is the author of ‘International Law and the Proliferation of Weapons of Mass Destruction (Oxford University Press, 2009); Interpreting the Nuclear Nonproliferation Treaty (Oxford University Press, 2011); and Iran’s Nuclear Program and International Law (Oxford University Press, 2016).
Image: Shutterstock/vectorfusionart
