NASA's 2022 fission power baseline: 40 kW, six tonnes, 120 volts

What the work claims

The document is a project overview, not a research result: as of October 2022, NASA and the Department of Energy intended to develop a space-qualified fission surface power flight unit before 2030, built by industry through a two-phase competition and bounded by a government reference concept with explicit key design characteristics: 40 kWe output at 120 Vdc, a 6,000 kg mass limit sized to fit on a lander, a radiation dose of no more than 5 rem per year above background at 1 km from the reactor, and the ability to operate either on the lander or after transport to a remote site.¹

The slide is careful about what is fixed and what is not. Under the key characteristics table sits one line: "Industry will determine the design." The 40 kW, six-tonne, 120 V figures are requirements levied on competitors, not descriptions of any existing machine. Weight the claim accordingly: this is a funded acquisition baseline, not a capability demonstration.

How the program was built to work

The mechanism here is institutional, and it matters as much as the reactor physics. NASA Glenn runs the project; the Department of Energy is the collaborating agency, with Idaho National Lab managing the industry contracts and Los Alamos providing reactor expertise. NASA's side of the house develops a government reference concept and runs two parallel maturation lines, one for power conversion and one for nuclear technology, so that the eventual flight design inherits demonstrated components rather than inventing all of them at once.¹

Industry participation is split into two phases. Phase 1 put three contractor teams on one-year contracts for initial designs: IX, a pairing of Intuitive Machines and X-Energy with Maxar and Boeing; Lockheed Martin with BWXT and Creare; and Westinghouse with Aerojet Rocketdyne. Their deliverables are design documents, requirements, schedule, and cost estimates that feed Phase 2, which is defined as a separate, open, competitive procurement producing a qualification unit and a flight unit.¹

The physics motivation on the first content slide is the standard one, stated plainly: the lunar night is 14.5 Earth days long, Mars has recurring planet-wide dust storms that can last for weeks, and a fission system designed for a lunar demonstration is positioned as directly applicable to human Mars exploration.¹

Two pieces of arithmetic fall directly out of the requirements table, and both are worth making explicit because the source does not. Dividing the 40 kWe output by the 6,000 kg mass limit gives roughly 6.7 W per kg at the complete-flight-unit level; that is a design-goal bound on a mass limit, not a measured specific power, since as of this document no flight unit exists and industry has not yet determined the design. Dividing 40,000 W by the 120 Vdc bus gives a full-output current of roughly 330 A, which sets the scale of every contactor, cable, and connector in the surface distribution system the reactor is meant to anchor. The 5 rem per year figure is about 50 mSv per year at 1 km, using the exact conversion of 1 rem to 10 mSv.

The strongest case for taking the baseline seriously

The steelman rests on requirement stability. NASA's Glenn Research Center feature of January 2024, published sixteen months after this presentation and after the Phase 1 designs came in, repeats the same core numbers: the reactor should stay under six metric tons and produce 40 kW of electrical power, and NASA's program office confirmed three 2022 contracts of $5 million each were awarded for the initial designs.² When a requirements set survives contact with three competing preliminary designs, it stops being a placeholder and starts being a procurement anchor. That is the single most encouraging fact in either document.

The second pillar is the structure. Separating reactor expertise (Los Alamos) from contract management (Idaho National Lab) from flight systems integration (NASA Glenn) puts each piece with the institution that already owns the relevant licensing, test infrastructure, and workforce. And keeping Phase 2 open and competitive rather than sole-sourced to a Phase 1 winner preserves price and design pressure all the way to the flight unit.¹

The third pillar is the honest treatment of conversion. By funding Brayton power converter development as its own line, the project acknowledges that the weakest link in a space fission system is usually the thermal-to-electric conversion chain, not the reactor itself; NASA's 2024 feature names Rolls Royce North American Technologies, Brayton Energy, and General Electric as converter contractors, which shows the maturation line from the 2022 slide converted into real contracts.²

Where a skeptic should push

Start with schedule, because the source itself is already out of date on its own most quotable commitment. The 2022 slide says "before 2030." NASA's January 2024 feature says open solicitation for Phase 2 is planned for 2025 and that the target date for delivering a reactor to the launch pad is the early 2030s, with a one-year lunar demonstration followed by nine operational years.² A flight unit "before 2030" quietly became launch-pad delivery "early 2030s" within sixteen months. Nor did Phase 1 run to plan: the 2024 feature says NASA extended the three Phase 1 contracts to gather more information before setting Phase 2 requirements, rather than proceeding on the one-year cadence the 2022 slide describes.² None of this is scandalous; all of it is schedule reality that the 2022 document, read alone, would hide from you.

The most load-bearing technical assumption is the coupling between the 6,000 kg mass limit and the 5 rem per year dose at 1 km. For a surface reactor, shielding is typically the mass driver, and the dose constraint at such short standoff forces either heavy shadow shielding, clever siting, or both. The deck does not say how the reference concept splits the mass budget between reactor, shield, and conversion; until Phase 2 designs freeze, the 6.7 W/kg figure computed above is an upper bound that competes against the dose requirement, not an achieved specific power. If shielding eats the margin, net electrical output or transportability is what gives.

Finally, TRL honesty: nothing in either document is hardware at flight maturity. The 2022 deck is an acquisition plan; the 2024 feature reports design studies, contract extensions, and converter development. Demonstrated maturity as of these sources is at the preliminary-design and component-maturation level. Anyone citing this program as evidence of a working lunar reactor is laundering a plan into a capability.

What the 2022 baseline commits lunar power to

The non-obvious implication hides in the voltage. A 40 kW system on a 120 Vdc bus runs near 330 A at full output, which is low-voltage, high-current distribution on a world with abrasive conductive dust and no established connector standard. Every habitat, rover, and ISRU plant that wants FSP power must either design to a 120 V-class bus or carry its own conversion stage; the requirements table on the project overview slide is therefore quietly also the first draft of the lunar surface grid's electrical interface.¹

The genuine threat is interface instability. The Phase 1 extension and the open Phase 2 solicitation mean the numbers on the 2022 slide are still moving parts; a vendor that hard-codes 120 Vdc and 40 kW into flight hardware now is betting on a requirements set that NASA itself has not yet frozen. The same openness that produced healthy design variety, as the 2024 feature puts it, also means downstream power consumers have no firm contract to design against yet.²

The opportunity is equally specific: because Phase 2 is defined as an open, competitive procurement with a published reference concept, it is one of the few places in space power where a new entrant can design to an actually funded government interface rather than to a white paper. The requirements table and the two-phase structure on the project overview slides are, in effect, a public request for the lunar grid's anchor hardware.¹

The bottom line

Calibrated judgment, bounded to what these two primary documents support: the commitment is real. Money moved (three $5 million design contracts, named converter awards), institutions with actual nuclear credentials are in place, and the 40 kW, six-tonne requirement survived its first contact with competing designs. I would put high confidence on NASA and DOE continuing to fund fission surface power through Phase 2. I would put moderate confidence on a reactor reaching the launch pad in the early 2030s, and low confidence on a lunar demonstration before the end of this decade, since even the 2024 plan lands hardware on the Moon only after the early-2030s delivery. What would confirm the program: a Phase 2 award with a frozen interface specification holding the 40 kW and mass limits. What would break it: shielding-to-dose trades that force net output below the demonstration threshold, or another schedule slide moving right without a Phase 2 award behind it.