U.S. Department of Energy (DOE) Authorization Is Not a Commercialization Pathway

Intended audience: Investors and family offices evaluating advanced reactor companies pursuing DOE authorization, and the founders raising capital on the strength of DOE milestones.

Executive Summary

A DOE authorization lets a developer build and operate a demonstration reactor. It does not let that developer sell electricity commercially, and any revenue generated during a DOE demonstration is incidental to the demonstration's purpose, not a business model. I've seen investors treat a DOE milestone as a proof point on the path to return, when in reality it's the start of a second, separate, and expensive regulatory process with the U.S. Nuclear Regulatory Commission (NRC), one that a developer can fail to plan or budget for even after a technically successful demonstration. This post lays out what DOE authorization actually is, what it isn't, and five questions investors should ask before treating a DOE milestone as evidence of a path to commercial revenue. For the technical mechanics of what DOE-authorized work can and can't carry forward into an NRC application, see my earlier post, "From DOE Authorization to NRC Licensing."

What a DOE Authorization Actually Is

A DOE authorization permits a developer to build and operate a demonstration reactor, typically at a DOE site, under DOE's own safety framework rather than the NRC's. The current wave of these demonstrations moves fast, intentionally so, because DOE's mandate is technology demonstration, not commercial licensing.

The Crusoe and Aalo Atomics partnership, announced July 30, 2026, is a useful, current example, and a good one, since Aalo's own public materials get this distinction right. Aalo will power a Crusoe Spark data center at Idaho National Laboratory in 2027 as a proof of concept, under DOE's Reactor Pilot Program. That's real, and it's a meaningful technical milestone. But deploying the commercial Aalo Pod at Crusoe's actual data centers, the 2029 target both companies have described, still requires NRC licensing, an application Aalo has not yet publicly filed. The DOE demonstration and the NRC commercial deployment are two different regulatory questions, answered by two different agencies, on two different timelines.

You Can't Read the Safety Case Yourself

Here's a due-diligence gap specific to DOE-authorized programs that's easy to miss. An NRC applicant's safety analysis report, and most of the documentation behind it, becomes part of the public record through the NRC's ADAMS database. An investor, or a competitor, or a curious member of the public, can go read it.

DOE's process doesn't work that way. The core safety document for a DOE nuclear facility is called a Documented Safety Analysis (DSA), and multiple sources describe DOE's review of Aalo's DSA for Aalo-X as comparable in rigor to an NRC Final Safety Analysis Report, DOE-Idaho approved it on April 30, 2026 after a genuinely serious review. But the DSA itself isn't published anywhere. What's public is a press release announcing that DOE approved it, not the underlying safety basis document. An investor evaluating a DOE-track company is, in a real sense, taking the rigor of that review on faith, because unlike an NRC applicant's SAR, there's no ADAMS-equivalent repository where a DOE DSA can be independently pulled and read. That's not a knock on DOE's process. It's a real difference in what an outside investor can actually verify, and it's worth knowing before treating "DOE approved" as equivalent to "I can independently see why."

A Demonstration Has a Lifespan, and an End-of-Life Cost

An operating DOE demonstration typically runs for a matter of months to a few years, not decades. When it ends, the developer faces decommissioning and used fuel disposition, real costs, even for a small demonstration unit, and even for fuel that was only lightly used. Any electricity sales that happen during the demonstration period are incidental to the demonstration's actual purpose: generating the data and operational experience the developer needs. A demonstration is not a revenue-generating asset with a long operating life. It's a data-collection exercise with a defined end date and a bill due at that end date.

Even in a well-documented, current example like the Aalo-Crusoe partnership, this is not public information. Neither company has disclosed a specific planned operating lifetime for Aalo-X, whether Crusoe will earn any revenue from electricity or compute output during the demonstration period, or whether the Crusoe Spark unit itself will need to be decommissioned or relocated once the demonstration concludes. That's not a criticism of either company. It's exactly the kind of information a real investor in this specific deal would need to ask for directly, because it isn't in the press release.

The Data Acceptance Problem

Here's the assumption I see trip developers up most often: assuming that data collected during a DOE demonstration will automatically be usable in a future NRC application. It won't be, not automatically. Unless the test plan that generated that data was reviewed and approved by NRC staff in advance, and unless the data was collected under a quality assurance program that meets NRC expectations, the NRC is under no obligation to accept it, and in practice, probably won't. DOE's acceptance of the data for its own purposes has no bearing on whether the NRC will accept it for licensing purposes. These are two separate agencies with two separate QA standards, and passing one doesn't mean automatically passing the other.

I've worked with a developer that ran directly into this. They assumed NRC acceptance of their DOE-collected data was a given, discovering only later that without NRC-approved methodology and test plans and NRC-caliber QA behind the data collection, that data carried far less weight in an NRC review than they'd budgeted for, in both time and money.

What Happens After the Demonstration Ends

A successful DOE demonstration is genuinely useful for raising capital. It's real evidence a developer can build and operate what they designed, and it can help unlock the kind of $100 million-plus raises that fund the next phase. But that next phase is the NRC application itself, a substantial, separately budgeted, multi-year undertaking that starts largely from the demonstration's finish line, not from a position with the license already halfway won. A developer that treats the DOE demonstration as the finish line, rather than the jumping-off point for an entirely separate regulatory process, is setting investors up for a serious expectations gap.

What This Means for Investors

The core distinction is this: DOE can authorize you to build a demonstration. Only the NRC can authorize you to sell the power it makes, at commercial scale, on an ongoing basis. If a developer's roadmap ends at "DOE demonstration complete" without a funded, staffed, scoped plan for what comes next with the NRC, there is no near-term path to a return on that investment, only a very expensive, well-documented science project.

Five Questions to Ask Before Investing on the Strength of a DOE Milestone

1. Does this authorization include commercial electricity sales, or is any revenue incidental to a demonstration program? If the answer is the latter, don't model demonstration-period revenue into a return case. It isn't the business.

2. Was the test and data collection plan reviewed and approved by NRC staff in advance, or only by DOE? DOE acceptance of a test plan or its resulting data does not mean the NRC will accept it. Ask specifically whether NRC staff reviewed the test plan and whether data collection ran under an NRC-caliber quality assurance program.

3. What happens to the demonstration unit when the test program ends, who pays for decommissioning and used fuel disposition, and is that cost already budgeted? A demonstration has a defined end date and a real bill attached to it. Make sure that bill isn't a surprise buried past the current raise.

4. Is there a funded, scoped NRC application plan, or just an intention to "pursue NRC licensing eventually"? A roadmap slide that ends at "DOE demonstration" and picks back up at "commercial deployment" with nothing budgeted in between is not a plan, it's a gap.

5. What is the realistic timeline and cost from "DOE demonstration complete" to "NRC license in hand," and does the current raise or runway actually account for it? This is usually years and tens of millions of dollars, not a footnote. If the fundraising narrative doesn't reflect that, the narrative is incomplete.

Frequently Asked Questions

Q: Can a company sell electricity commercially under a DOE authorization?

A: Generally no. DOE authorization permits demonstration operation, and any electricity sales during that period are incidental to the demonstration's purpose. Commercial electricity sales at scale require a separate NRC license.

Q: If a developer collects test data under a DOE demonstration, will the NRC automatically accept that data in a future license application?

A: No. The NRC's acceptance of demonstration data depends on whether the test plan that generated it was reviewed and approved by NRC staff in advance, and whether the data was collected under a quality assurance program that meets NRC expectations. DOE's own acceptance of the data has no bearing on the NRC's evaluation.

Q: Can an investor read the actual safety analysis behind a DOE-authorized demonstration reactor, the way they could read an NRC applicant's safety analysis report?

A: Not directly. DOE's core safety document for a nuclear facility is called a Documented Safety Analysis (DSA), and DOE's review of it can be genuinely rigorous, comparable in scope to an NRC Final Safety Analysis Report by some accounts. But unlike an NRC application, which becomes part of the public record through the NRC's ADAMS database, DOE does not publish DSAs. What's public is typically a press release confirming DOE's approval, not the underlying document. Investors evaluating a DOE-track company should be aware they're relying on DOE's word for the review's rigor, not an independently verifiable public record.

Q: What happens to a demonstration reactor after the test program is complete?

A: The developer is responsible for decommissioning the unit and for disposition of any used nuclear fuel, even fuel that saw only limited use. Both carry real costs that should be budgeted for from the start of the demonstration program, not discovered afterward.

Q: Is a successful DOE demonstration a good fundraising milestone?

A: Yes, genuinely. It's real evidence a developer can build and safely operate what they designed, and it's historically helped developers unlock large follow-on raises. The risk isn't the milestone itself, it's investors and developers treating it as the finish line rather than the starting point for a separate, expensive NRC regulatory process.

Q: How does this relate to the DOE-to-NRC transition guidance you've written about before?

A: This post addresses the investor-facing question of whether DOE progress equals commercial readiness. For the technical mechanics of what DOE-authorized work can and can't carry forward into an actual NRC application, including QA pedigree requirements and the current draft interim staff guidance, see my earlier post, "From DOE Authorization to NRC Licensing."

Sarah Gibboney, P.E. is Founder & Principal Licensing Engineer of Gibboney Nuclear, PLLC, a nuclear licensing consultancy serving advanced reactor developers. She has 17 years of nuclear energy experience, including co-authoring Construction Permit Applications for both ARDP awardees, TerraPower Natrium and X-energy Xe-100.