NASA and DOE Signed a 10-Year Space Nuclear Treaty. It Takes Effect One Day After the Old One Expires.

On October 8, 2026, at the White House Office of Science and Technology Policy's "Science: A New Golden Age" summit, NASA Administrator Jared Isaacman and Energy Secretary Chris Wright signed a memorandum of understanding titled "Accelerating American Leadership in Space Nuclear Power and Propulsion." It takes effect November 1, 2026 and runs for ten years unless revised or extended.

Pay attention to the date arithmetic, because it is the most informative thing in the document.

The MOU amends and supersedes two prior agreements: a radioisotope power systems MOU signed October 31, 2016 — which was set to expire October 31, 2026 — and a space fission systems MOU signed December 22, 2025. The new agreement becomes effective November 1. That is one day after the old one lapses.

This was not a vision statement drafted to fill a summit agenda. It was a bridge built to cover a gap that would have opened in three weeks. The 2016 agreement was the legal plumbing underneath every radioisotope power system NASA flies — the plutonium-238 heat sources that run Perseverance and Dragonfly. Without a successor, the interagency framework governing that supply chain would have gone dark on Halloween.

The second superseded agreement is more interesting still. The space fission systems MOU was signed in December 2025 and was replaced within roughly ten months. Agencies do not rewrite a ten-month-old agreement because the formatting was wrong. They rewrite it because the program underneath it changed shape.

What the MOU actually names

Two flight programs appear by name, with dates:

• Space Reactor-1 "Freedom" — a fission-powered spacecraft demonstrating nuclear electric propulsion, targeted to launch in December 2028, headed beyond Earth orbit toward Mars.
• Lunar Reactor-1 — the fission surface power system intended to keep a planned American lunar base operating through the two-week lunar night, with a launch-ready unit due by 2030.

A third item surfaces in the surrounding record: a Radioisotope Stirling Generator Prototype solicitation issued in September 2026 for lunar applications.

Existing interagency agreements between the two agencies remain in effect. The MOU sits on top of them as the governing framework, not a replacement for the task-level paperwork.

The division of labor is the part you should read twice

The MOU allocates responsibility roughly like this:

DOE owns: design, development, fabrication, fuel production as necessary, nuclear material authorization, and technical assistance.

NASA owns: space mission authority, requirements definition, system installation and testing, launch integration, operations — and the authority to "procure or partner on commercially designed and developed space nuclear systems."

If you build hardware in this sector, that split tells you who signs your contract. Anything inside the nuclear boundary — fuel forms, core components, anything requiring nuclear material authorization — routes through DOE and its national laboratories. Anything outside it — power conversion integration at the spacecraft level, radiators and heat rejection, power management and distribution, structures, avionics, launch integration — routes through NASA or a NASA prime.

Teams get this wrong constantly, and it costs a full capture cycle. A company with a Stirling convertor is selling to a different buyer than a company with a uranium fuel form, even though both show up at the same industry day and both think they are "in fission surface power."

Six years of scope growth, compressed into one agreement

The historical context matters for anyone sizing this market.

In 2020, NASA and DOE were preparing a solicitation for a 10-kilowatt fission surface power system that officials hoped could be launch-ready "as soon as 2027." The acquisition was structured as a DOE-led two-phase process: multiple Phase 1 preliminary design awards, then a single company selected to build the flight unit. An August 2020 industry day drew more than 180 participants, spanning nuclear firms like BWXT and General Atomics and aerospace firms like Blue Origin and Lockheed Martin. Officials used the word "partnering" heavily, because no single company had both the reactor expertise and the spaceflight expertise.

DOE issued the formal request for proposals in 2021. In June 2022, three design concept contracts worth $5 million each were awarded. In January 2025, Westinghouse was selected by NASA and DOE to continue developing its space microreactor design.

Today the Fission Surface Power project targets a 40-kilowatt class system — NASA describes it as enough to continuously run 30 households for ten years — with operational lunar deployment in the early 2030s. The lineage runs back further, to the Kilopower program and the 2018 KRUSTY experiment that demonstrated the heat-transfer physics.

Read those two paragraphs together and the trend is unmistakable: the power requirement quadrupled and the schedule slipped about three years. That is not a criticism of the program. It is the single most useful planning input available to anyone writing a proposal into this space. A 10 kW design study became a 40 kW hardware program, and the money moved accordingly — from $5 million concept studies toward fabrication, qualification, and fuel.

Where an actual applicant enters

Here is the hard part, and we want to be blunt about it: an MOU is not a funding opportunity. It appropriates nothing, obligates nothing, and will never appear on Grants.gov. If you are watching Grants.gov for the effects of this agreement, you will see nothing and conclude wrongly that nothing happened.

The practical entry points for universities and small businesses are the mechanisms that already exist:

• NASA SBIR/STTR. NASA has restructured its program into a rolling Broad Agency Announcement with continuously released appendices, Phase I at $225,000 and Phase II at $1.275 million. Space nuclear subsystems — thermal management, radiation-tolerant electronics, autonomous control, power conversion — are recurring topic territory. See our analysis of the NASA SBIR/STTR BAA restructuring.
• DOE Office of Science open call. DE-FOA-0003665 remains open for rolling submissions through September 30, 2027, with roughly $400 million behind it and a pre-application review panel calendar that rewards early contact. Our breakdown of the open call covers how the panel timing works.
• DOE national laboratory announcements and lab calls, where university teams partner rather than prime.
• Irradiation and testing access through DOE user facilities, which is often the binding constraint on a fuel or materials proposal — not the research dollars.
• NASA ROSES for the science payloads that radioisotope and fission power enable.

The gate nobody budgets for

If you intend to put nuclear material on a launch vehicle, your schedule is not governed by your engineering. It is governed by nuclear launch safety approval — the tiered interagency launch authorization process, including the safety analysis that precedes it. Budget twelve to twenty-four months and a substantial documentation burden, and never promise a flight date in a proposal without showing the authorization path alongside it. A reviewer who has lived through this process will discount a 2028 flight claim that does not address it.

How much durability should you assume?

Be calibrated. The agreement this MOU replaced was signed in December 2025 and lasted about ten months. The one before that lasted its full ten-year term. A ten-year framework signed at a political summit, with named flight dates in 2028 and 2030 and outyear funding dependent on appropriations Congress has not passed, is a strong demand signal and a weak guarantee.

What it genuinely does is remove ambiguity about who buys what for the next decade, and it removes the October 31 expiry cliff that would otherwise have frozen radioisotope work. For a small business deciding whether to spend internal R&D on a space-qualified Stirling convertor or a radiation-hardened power controller, that clarity is worth more than another press release with a bigger number in it.

Watch SAM.gov and NASA's procurement forecast, not Grants.gov. The FY2027 procurement cycle is where a December 2028 launch gets its hardware, and that cycle is starting now.