nuclear-reactor

Fission surface reactor

Subsystem Hard import Seed import
TRL Mars
6 / 9
Energy intensity
—
Required by
1
Requires
2

Fission of U-235 in a controlled reactor releases ~80 TJ per kg of fuel — 2.4 million times the energy density of coal. NASA's Kilopower (KRUSTY) demonstrated a 1–10 kWe Mars-class fission surface power system in 2018 using a sodium-heat-pipe-cooled core, highly enriched uranium fuel, and free-piston Stirling engines. Multi-megawatt micro-modular reactors are the next step for a settlement of thousands.

Last reviewed: 2026-06-08

Governing equations

Single fission event. Two fission products (FPs), 2–3 prompt neutrons that sustain the chain reaction, and ~200 MeV of energy as kinetic energy of fragments + decay heat + neutrinos. [1]

Effective multiplication factor. k_eff = 1 is criticality (steady-state); >1 is supercritical (power rises); <1 is subcritical (power decays). Reactor control = holding k_eff at 1. [1]

Decay heat as a fraction of operating power, t seconds after shutdown. ~7 % at shutdown, 1 % at 1 hour, 0.4 % at 1 day. Mars reactors must reject this heat passively or the core melts even after scram. [1]

Radiator area for vacuum heat rejection. Mars near-vacuum (~600 Pa CO₂) means radiation, not convection, is the only practical sink. Hot-side T ≥ 800 K is mandatory to keep radiator mass tractable. [2]

Key constants & quantities

Symbol Value Units Conditions Description
E_fission 200 ±5% MeV / fission — Energy released per U-235 fission. ~3.2 × 10⁻¹¹ J per event. Of this, ~190 MeV is recoverable as heat; ~10 MeV escapes as neutrinos.[1]
ED_U235 80 TJ / kg U-235 — Energy density of U-235 fuel. 2.4 × 10⁶ × coal, 3 × 10⁶ × diesel. Drives the entire mass-budget case for nuclear on Mars.[1]
P_KRUSTY 1–10 kWe (electrical) — Output range of Kilopower Reactor Using Stirling TechnologY — the Mars surface fission reference design. Scalable in 1, 5, 10 kWe modules.[2]
m_KRUSTY 1,500 ±10% kg (system) — Total system mass for a 10 kWe Kilopower unit including reactor, shield, Stirling, radiator. ~150 kg/kWe — high vs terrestrial, low vs PV + battery for Mars.[3]
η_Stirling 25–35 % — Thermal-to-electrical conversion efficiency of free-piston Stirling engines at hot-side 1100 K, cold-side 400 K. Competitive with steam Rankine at these scales.[2]
τ_design 15 years — KRUSTY design life. 130 000 hours of continuous operation without refueling — limited by control-rod drive wear and Stirling regenerator fatigue, not fuel burnup.[3]
σ_shield 4 m regolith equivalent — Regolith berm thickness to reduce dose at a 100-m-distant habitat to < 5 mSv/year. Combines reactor shadow shield (tungsten + lithium hydride) with regolith berming.[4]

Operating envelope

ParameterRangeUnitsSource
Core temperature 600 – 950 °C [3]
Stirling hot-side 800 – 1100 K [2]
Radiator T (hot-side) 400 – 600 K [2]
Burnup at EOL 1 – 5 % U-235 consumed [3]
Enrichment (HEU) 19.75 – 93 % U-235 [4]

Mass balance

Basis: 10 kWe Kilopower system, 15-year design life

Inputs

HEU fuel (UO₂) 28 kg [3]
BeO neutron reflector 75 kg [3]
Sodium (heat-pipe working fluid) 1.2 kg [2]
  • HEU fuel (UO₂): 93% enriched U-235 in a single solid monolith. Total fuel burnup ~ 2% over life.
  • BeO neutron reflector: Reflects escaping neutrons back into core; allows smaller fissile mass.
  • Sodium (heat-pipe working fluid): Eight heat pipes carry reactor heat to Stirling hot-side.

Outputs

Electrical energy (15 yr) 1,314,000 kWh [3]
Spent fuel + activated structure 105 kg [3]
Waste heat (radiated) 26,280 MWh (thermal, lifetime) [2]
  • Electrical energy (15 yr): 10 kWe × 8760 h/yr × 15 yr × 1.0 capacity factor.
  • Spent fuel + activated structure: Fission products, activated reflector, residual fuel. Encapsulated for surface burial or return.
  • Waste heat (radiated): 40 kW thermal × 8760 h × 15 yr — radiated to Mars sky from the cone radiator.
TRL · Earth
9/ 9
TRL · Mars
6/ 9
Terrestrial fission is TRL 9 (commercial since 1957). KRUSTY hit TRL 6 with the 2018 Nevada test — first U.S. new-design reactor demonstrated since 1977, full-power criticality + Stirling generation. Fission Surface Power (FSP) program funded by NASA + DOE in 2022 to deliver 40 kWe Mars-deployable units by 2030. No fission reactor has yet operated on Mars; the Mars TRL is set by the Earth-tested KRUSTY architecture being directly transferable.[2]
Energy budget
0.667 kWhe / g U-235 consumed + 2 kWhth [1]

1 g U-235 fully fissioned = 24 MWh = 24 000 kWh thermal. At 30 % Stirling efficiency, ~8 000 kWh electrical per gram. Per kWh, this works out to ~0.125 mg U-235 — the entire 15-year life of a Kilopower system fits in a thimble of fissioned uranium.

Variants & trade-offs

Kilopower (KRUSTY-class)

[2]

Highly enriched UMo monolith fuel, sodium heat pipes, free-piston Stirling, BeO reflector with B₄C control rod. 1–10 kWe per module. The NASA-tested Mars reference.

Electrical output
1–10 kWe
Core temperature
800–850 °C
Stack lifetime
100000–130000 h
Materials: UMo HEU monolith fuel · BeO reflector · B₄C control rod · Na heat pipes (Hastelloy) · Stirling regenerator (stainless mesh)
  • Mars-tested architecture (KRUSTY 2018) — only NASA-funded Mars fission to TRL 6
  • Passive shutdown — control rod gravity-falls into core on power loss
  • Modular: cluster 4× 10 kWe units for 40 kWe surface power
  • No coolant pumps — heat pipes are passive
  • HEU (93% U-235) is a non-proliferation challenge — politically constrained
  • Stirling pistons are the only moving parts but also the lifetime-limiter
  • Scales poorly above 10 kWe — heat-pipe limitations
  • Single-string fuel monolith: no in-flight refueling

Fission Surface Power 40 kWe (NASA/DOE 2030 target)

[4]

Next-generation Mars reactor under contract since 2022. Multiple teams (Lockheed/BWX, IX/Maxar, Westinghouse) competing for the 40 kWe, 10-year, < 6 t mass, < 4×6 m footprint spec.

Electrical output
40–40 kWe
Mass cap
6–6 t
Stack lifetime
87600–87600 h (10 yr design)
Materials: Low-enriched UMo or UN fuel (< 20% U-235) · Brayton or Stirling power conversion · Heat-pipe or liquid-metal primary loop
  • Designed specifically for Mars surface deployment
  • LEU fuel avoids HEU non-proliferation issues
  • Sized for a 4-person + early base load (life support + ISRU)
  • Still pre-flight — first launch is 2030+ in the optimistic plan
  • TRL 4–5 for integrated 40 kWe operation
  • Competing architectures — final design not yet downselected

Micro-modular reactor (terrestrial Gen-IV)

[1]

1–100 MWe land-based reactors (NuScale, BWXT, Oklo, X-energy). Heavier but vastly more powerful; relevant once a Mars colony scales past ~100 people and needs MWe-class power for industry.

Electrical output
1–100 MWe
Operating temperature
285–950 °C (variant-dependent)
Materials: UO₂ pellet fuel (LEU) · Zircaloy cladding · Water or molten-salt or gas coolant
  • Megawatt-class output enables heavy industry (electric arc furnace, fertilizer, ISRU at scale)
  • Multiple Earth-side designs near commercial readiness (NuScale licensed 2023)
  • Long fuel cycle (5–10 yr between refueling)
  • > 100 t system mass — Mars transit unrealistic until launch costs drop further
  • Designed for terrestrial regulatory + grid integration, not autonomous Mars surface
  • Conventional water coolant needs makeup water — adds Mars dependency

Failure modes

Mode Cause Detection Mitigation
Loss of coolant accident (LOCA)[2] Heat-pipe rupture, primary-loop leak, or radiator damage interrupts heat removal from the core. Core T rises, control system signals coolant flow / pressure loss, automatic scram. Heat-pipe redundancy (8 pipes in Kilopower — losing 2 is survivable); shadow shield doubles as containment; decay heat removed passively to regolith via conduction.
Control rod stuck or failed[3] Mechanical jam, actuator failure, or B₄C poisoning of the rod itself. Reactor power doesn't respond to commanded rod motion. Multiple independent rods; gravity-fall safety design (rod weight inserts on power loss); soluble boron injection as backup.
Decay heat removal failure[1] After shutdown, fission products continue producing ~7% of operating heat that decays over hours. Loss of all heat removal melts the core in minutes to hours. Core T rises after shutdown despite zero fission. Passive radiator that doesn't require pumps; conductive sink to surrounding regolith; final fallback is core melting into safe geometry (Inherent safety design).
Fuel cladding breach[1] Thermal cycling fatigue, fission-gas pressure buildup, or manufacturing defect. Activity in coolant loop (Earth) or noble-gas release outside shield. Cladding integrity monitoring; conservative fuel temperature limits; redundant containment shield around the core.
Radiation damage to power conversion[2] Gamma + neutron flux degrades Stirling regenerator alloys and seal polymers over years. Conversion efficiency declines below design point. Shadow shield between core and Stirling; radiation-tolerant materials; redundant convertors with replaceable warm-side seals.
Dust accumulation on radiator[4] Mars dust storms deposit µm-scale fines on radiator surfaces, reducing emissivity and effective heat rejection. Radiator T rises at constant power; core T trends up. Vertical or near-vertical radiator orientation; electrostatic dust shedding; periodic mechanical wipe; design margin for 30 % dust derate.

Mars adjustments

Radiator sizing in near-vacuum[2]

Impact: No convection — heat rejection is radiative only. A 40 kW thermal radiator at 400 K cold-side needs ~80 m² of surface; raising hot-side to 600 K cuts this 5× via the T⁴ relation.

Mitigation: High-temperature primary loop (Brayton or high-T Stirling) to push radiator T above 500 K. Conical or fan radiator geometries minimize self-view-factor losses.

Shielding via regolith berming[4]

Impact: Shipping enough tungsten or LiH to shield a fission core for 15 years of crew exposure costs hundreds of kilos per kWe. Mars regolith at the deployment site is free shielding material.

Mitigation: Bury or berm the reactor under 3–4 m of regolith using site-prep equipment. Shadow shield (~150 kg in Kilopower) handles the unshielded direction toward the habitat. Combined reduces crew dose below 5 mSv/year at 100 m.

No regulatory framework on Mars[4]

Impact: Earth nuclear deployments require NRC + IAEA licensing, which depends on national jurisdiction. Mars has no equivalent — but the Outer Space Treaty (1967) makes the launching state responsible.

Mitigation: NASA/DOE-led launch authorization under U.S. NEPA + DOE-Order-5480.6 (space nuclear). Settlement-era Mars will need a domestic framework — likely modeled on shipping conventions.

Dust-storm power continuity[5]

Impact: Sol-long or weeks-long dust storms can cut PV output to 10% of nominal. Fission output is unaffected — this is the architectural reason for nuclear, not solar, as the foundational Mars power source.

Mitigation: Nuclear baseload + PV as supplementary daytime augmentation. Storm derates of nuclear are limited to ~10–15% radiator-fouling impact, vs 90% for PV.

Decommissioning + waste management[5]

Impact: A spent Kilopower core is ~100 kg of activated metal + fission products at end of life. No Earth-equivalent geological repository on Mars; bringing it home is launch-mass expensive.

Mitigation: Mars Direct architecture: encapsulate spent core in welded steel cask, bury at 50+ m depth in a stable geological feature. Long-half-life isotopes decay over geological timescales; Mars geology is more seismically inert than Earth.

Alternatives & substitutes

Photovoltaic array + battery storage[5]

  • No regulatory or non-proliferation concerns
  • Modular and incrementally deployable
  • Mature TRL 9 for both PV and batteries
  • Mars solar constant 43% of Earth; cosine + dust + season cuts further
  • Dust storms can derate to 10% for weeks (Mariner 9, MER missions)
  • 50–100 m² PV + 1000 kWh battery per kWe steady — orders-of-magnitude more launch mass than fission
  • Cannot supply night-side or polar-winter operations alone

When preferred: Early scouting missions; supplementary daytime power to nuclear; equatorial-only ops with short outage tolerance.

Radioisotope thermoelectric generator (RTG)[1]

  • Decades of flight heritage (Voyager, Curiosity, Perseverance)
  • No moving parts
  • Mars-environment-proven
  • ~100 W per RTG — three orders of magnitude below kilopower needs
  • Pu-238 supply is constrained by U.S. production (~ 1.5 kg/year)
  • Efficiency 6% — most decay heat wasted

When preferred: Instrument-scale power (sensors, rovers); never settlement-scale.

Inputs: Nuclear fuel (U-235) Built from: Stainless steel stock Outputs: Electricity

Requires

Inputs

Built from

Required by

References

  1. International Atomic Energy Agency (2022). Nuclear Reactor Physics and Engineering (compiled IAEA technical reports). IAEA, Vienna. — Reference compendium for reactor physics: fission energetics, criticality, decay heat (Way-Wigner).
  2. Mason, L., Gibson, M., Poston, D., Briggs, M., Sanzi, J., & Bell, J. (2018). A Small Fission Power System for NASA Exploration: KRUSTY Test Results. Nuclear and Emerging Technologies for Space (NETS) Conference, Las Vegas. NASA/TM-2018-219782. — KRUSTY full-power test 2018; Mars surface fission TRL 6 demonstration.
  3. Gibson, M. A., Oleson, S. R., Poston, D. I., & McClure, P. (2017). NASA's Kilopower Reactor Development and the Path to Higher Power Missions. IEEE Aerospace Conference, Big Sky, MT. doi:10.1109/AERO.2017.7943946 — Kilopower architecture, mass budget, Stirling integration, 15-year design life.
  4. Poston, D. I., McClure, P. R., Dixon, D. D., Gibson, M. A., Mason, L. S., & Sanchez, R. G. (2020). Experimental Demonstration of a Reactor Core Heated by Heat Pipes. Nuclear Technology, 206(sup1), S15-S30. doi:10.1080/00295450.2020.1722554 — KRUSTY follow-on; Mars Fission Surface Power roadmap; shielding architecture.
  5. Zubrin, R., & Wagner, R. (1996). The Case for Mars: The Plan to Settle the Red Planet and Why We Must. Free Press, New York. — Mars Direct mission architecture, in-situ propellant production, water electrolysis context.