Permanent lunar power should be fission-first, solar-secondary, storage-heavy, and repairable for centuries. Solar alone is too fragile for a 14-day night cycle; RTGs are too low-power; batteries alone are too mass-inefficient; and any credible 1000-year architecture must treat power generation as a modular, replaceable industrial utility, not a single asset.
Bottom line
- Near-term base load: a 10 kWe-class Kilopower-derived reactor is the most mature continuous-power concept in the available data, with NASA describing Kilopower as capable of 1 to 10 kWe continuously for 10 years or more.
- Forward lunar base target: NASA’s Fission Surface Power (FSP) effort is explicitly aimed at a 10-kilowatt-class lunar demonstration and later 40–100 kW systems, with a 2026 source stating the updated effort now seeks at least 100 kWe using a closed Brayton cycle[7].
- Night survival: for a facility that must ride through a full lunar night, regenerative fuel cells (RFCs/RFCS) are the strongest non-nuclear storage option in the data, with NASA and ESA both treating them as viable for lunar night power[3][8].
- Strategic conclusion: a permanent autonomous lunar settlement should use fission for guaranteed baseload, solar for daytime surplus, and chemical storage or thermal buffering for night/contingency, with solar and storage components designed for replacement, not indefinite untouched operation.
1) Solar arrays: performance and degradation over decades
Solar arrays on the Moon are attractive because the lunar day is long and sunlight is strong, but they face three long-term killers: radiation, micrometeoroids, and dust. The major planning problem is not one lunar night; it is maintaining output for decades without human refurbishment.
Key implications:
- Array degradation is cumulative, so a system sized only to initial output will underperform later.
- Dust accumulation is a major risk because lunar regolith is abrasive and electrostatically mobile.
- Mechanical deployment systems and cable harnesses become lifetime bottlenecks, not just the cells themselves.
A lunar base should assume solar output falls materially over time and therefore oversize arrays and inverters from day one. For a 1000-year horizon, the correct model is replaceable photovoltaic fields, not “permanent” panels.
Practical design rule:
- Build solar as a consumable industrial asset
- Overbuild generation to cover:
- end-of-life degradation
- seasonal geometry variation at the site
- dust losses
- storage charging margins
- emergency redundancy
2) Nuclear fission reactors: the only credible continuous baseload
### Kilopower
NASA describes Kilopower as a small, simple fission concept for long-duration, sun-independent power. NASA states it can produce 1 to 10 kilowatts of electrical power continuously for 10 years or more. The reactor uses a solid uranium-235 core with passive sodium heat pipes and Stirling conversion in the prototype power system.
Mission value:
- Continuous output
- No dependence on sunlight
- High reliability for autonomous operation
- Small mass relative to power continuity
Limitation:
- Kilopower-class output is useful for outposts, not large industrial bases.
### Fission Surface Power (FSP)
NASA’s lunar FSP effort expands Kilopower toward a real surface base power system. The cited NASA material says the project is pursuing a 10-kilowatt-class lunar demonstration in the late 2020s[7]. A 2026 source reports the updated program now targets at least 100 kilowatts of electrical power for lunar use.
Mission value:
- Makes sense for:
- habitat life support
- comms
- ISRU plant operation
- thermal management
- charging fleet vehicles and energy stores
- emergency survival margin
Why fission matters for a 1000-year archive site:
- It is the only near-term option that can provide predictable, weather-independent, night-independent power with manageable mass.
- It reduces dependence on fragile supply chains for daily life support.
- It can anchor the energy system while solar, storage, and ISRU systems cycle around it.
3) RTGs: useful for sensors, not for settlements
RTGs are excellent for low-power, ultra-long-duration applications, but they are fundamentally the wrong class for a lunar facility meant to support people, machines, and industry.
Limitations:
- Too little power for habitat loads, drilling, oxygen production, metals processing, or rover fleets.
- Thermal efficiency is low, so much of the nuclear energy becomes waste heat rather than electricity.
- Isotope supply is finite and strategically scarce.
- Power scales badly: adding enough RTGs for a base becomes mass-heavy and impractical.
Correct role:
- RTGs are suitable for:
- remote instruments
- survival beacons
- deep-shadow sensors
- small autonomous science packages
Incorrect role:
- primary grid power
- night-time habitat support
- ISRU industrial loads
For a civilization backup site, RTGs should be treated as auxiliary watchdog power, not infrastructure power.
4) Energy storage for the 14-day lunar night
The lunar night is the central storage problem. NASA’s RFC system reference gives a concrete design case: 10 kW of constant power during lunar night, with the lunar night defined as 200 hours and lunar day as 500 hours, totaling 700 hours per cycle.
NASA and ESA both identify regenerative fuel cells as a strong solution because they separate power output from stored energy[2][3][4][8]. ESA describes the process plainly: daytime solar power splits water into hydrogen and oxygen, which are stored and later recombined in a fuel cell at night[3]. ESA also reported demonstrated lunar day-night profiles and technology readiness of TRL 5 for fuel cell and electrolyzer stacks[3].
### Regenerative fuel cells
The strongest storage numbers in the retrieved