The TRL 5 roadmap for megawatt-class nuclear electric propulsion

What the work claims

The TMP claims that megawatt-class nuclear electric propulsion is technically viable for a 2030s Mars mission, but that most key NEP technologies currently sit at or below TRL 4 at the component level, and lower at the subsystem and system level.¹ The plan targets reaching TRL 5 (component validation in a relevant environment) by the mid-to-late 2020s, which would position NASA to make a flight development commitment in time for a 2030s mission window.

The plan is explicit that the advancement degree of difficulty (AD2) is high enough to support multiple development paths, meaning there is more than one technically defensible route from the current state to TRL 5, and the choice between them involves cost, schedule, and risk trades that the TMP does not fully resolve. The document draws on a NASEM (National Academies) critical review that found a lack of coordinated research and development for TRL 5 megawatt-electric-class hardware, and a NASA Engineering and Safety Center (NESC) review that found most NEP technologies at or below TRL 4 component level.¹

How it works

Nuclear electric propulsion uses a fission reactor to generate heat, converts that heat to electricity (typically through a Brayton or Stirling cycle, or thermoelectric converters), and then uses that electricity to power electric thrusters. The thrusters ionize a propellant (typically xenon, krypton, or argon) and accelerate the ions through electric and magnetic fields to produce thrust. The key advantage is specific impulse: NEP thrusters achieve thousands of seconds of specific impulse, compared to 360 to 460 seconds for chemical propulsion.¹ Higher specific impulse means less propellant mass for a given velocity change, which is the fundamental trade for deep space missions.

The TMP organizes the technology base into the major subsystems that must mature together: the reactor (fuel, core, shielding, heat transport), the power conversion (Brayton, Stirling, or thermoelectric), the power management and distribution (PMAD), the electric thrusters (Hall effect, gridded ion, or magnetoplasmadynamic), and the thermal management (radiators, heat rejection). Each subsystem has its own TRL, and the system-level TRL is limited by the lowest component that must integrate with the others.

The advancement degree of difficulty (AD2) is a NASA metric that estimates how hard it is to move a technology from its current TRL to the next. The TMP finds AD2 high enough that multiple development paths are defensible, meaning the technology could mature through different sequences of reactor, conversion, and thruster choices, each with its own cost and risk profile. This is not a plan that picks one architecture. It is a plan that maps the trade space.

The strongest case

The strongest case for the TMP's approach is its TRL honesty. The NASEM critical review found a lack of coordinated R&D for TRL 5 megawatt-electric-class hardware, and the NESC review found most NEP technologies at or below TRL 4 at the component level. The TMP does not paper over this. It states the current state plainly and then maps what it would take to reach TRL 5 by the mid-to-late 2020s. That candor is more useful than an optimistic roadmap, because it tells a mission planner exactly what needs to be demonstrated and when.¹

The specific impulse advantage is real and physics-based. A 5,000-second specific impulse NEP system needs roughly an order of magnitude less propellant than a 450-second chemical system for the same velocity change. For a Mars mission, that propellant mass savings translates directly into lower launch mass, fewer launches, or more payload. The trade is power and time: NEP requires a megawatt-class reactor, which is heavy, and it produces low thrust, which means long transit times. The TMP evaluates these trades in the context of a 2030s Mars mission.

The historical investment base, while discontinuous, provides heritage. The SNAP program in the 1960s flew SNAP-10A, the Space Exploration Initiative in the 1990s funded reactor concepts, and Project Prometheus in the 2000s developed the Jupiter Icy Moons Orbiter concept with a multi-kilowatt reactor. None of these reached flight, but each left design heritage, test data, and a workforce that informs the TMP.¹

Where a skeptic should push

The TRL gap is the core skeptical point, and the TMP itself states it. Most NEP technologies are at or below TRL 4 at the component level, and lower at the subsystem and system level. TRL 4 means component or breadboard validation in a laboratory environment. TRL 5 requires component validation in a relevant environment. The gap between a bench-tested component and a relevant-environment-validated component is where most space nuclear technologies have historically stalled, because the relevant environment (vacuum, radiation, thermal, launch loads) is expensive to replicate at megawatt scale.¹

The timeline is aggressive. The TMP targets TRL 5 by the mid-to-late 2020s. The document was published in February 2026, which means the plan is asking for TRL 5 in roughly 2 to 4 years. Given that the NASEM review found a lack of coordinated R&D, and that the NESC review found the technology base below TRL 4, the question is whether the funding and test infrastructure can be stood up fast enough. Project Prometheus, which had comparable ambitions, was cancelled in 2006 after several years of investment without reaching TRL 5 on its reactor. The historical pattern is that space nuclear programs are started, partially funded, and then cancelled before flight hardware is built.

The reactor is the long pole. A megawatt-class space fission reactor has not been built or tested in the United States since the SNAP program. The fuel, the core, the shielding, and the heat transport loop all require ground testing in a relevant environment, which means a nuclear test facility capable of handling megawatt thermal output. The TMP does not specify whether that facility exists or must be built, and facility construction is a multi-year activity that does not appear in the TRL timeline.

Finally, the multiple development paths finding is a double-edged sword. It means the technology is flexible, but it also means no single path has been validated as the best. A program that cannot commit to one architecture risks spreading limited funding across multiple approaches, none of which reaches TRL 5 on schedule.

What it means for power in space

The non-obvious implication is that an NEP system is, fundamentally, a megawatt-class space power system that happens to use its electricity for thrust instead of for payload. The reactor, power conversion, PMAD, and radiator subsystems that the TMP maturation plan covers are the same subsystems needed for a megawatt-class lunar or Mars surface fission power system. The specific mechanism is the shared reactor and conversion technology base: a reactor designed to drive electric thrusters can, with different power conversion and heat rejection, drive surface loads. Investment in NEP maturation is investment in high-power space nuclear generation, regardless of whether the first application is propulsion or surface power.¹

The genuine threat is the cancellation cycle. The TMP itself documents the pattern: SNAP in the 1960s, SEI in the 1990s, Project Prometheus in the 2000s, all started and stopped without flight hardware. If the current NEP effort follows that cycle, the technology base will remain at TRL 4 or below, and the next Mars mission will use chemical propulsion with its propellant mass penalty. The specific mechanism is the dependence on sustained multi-year funding through a change of administration or budget priority, which has killed every prior US space nuclear program.

The opportunity is that the TMP, by being honest about the TRL state and the AD2, gives NASA a defensible basis for requesting the specific funding needed to close the gap. The NASEM and NESC reviews are independent validations of the current low readiness. A program manager can point to those reviews and the TMP to justify ground testing of a megawatt reactor and its power conversion, which is the single highest-value investment for both NEP and surface fission power. If the test infrastructure is funded, the TRL 5 target is credible. If it is not, the TMP is a well-documented plan that cannot execute.

The bottom line

What is solid is the assessment: the TMP, backed by NASEM and NESC reviews, establishes that megawatt-class NEP is at TRL 4 or below at the component level and lower at the system level, with high advancement difficulty. What is not established is any path to TRL 5 that does not require sustained funding and a ground test facility that may not yet exist. Confidence is moderate that the technology could reach TRL 5 by the late 2020s if the reactor and power conversion test infrastructure is funded now, and low that it will do so without that infrastructure. The reading would be strengthened by a funded ground test program for a megawatt-class reactor and its power conversion at relevant scale. It would be undercut by a repeat of the Prometheus pattern: several years of investment followed by cancellation before TRL 5, which is the historical default for US space nuclear programs.