Good luck colonizing the Moon without nuclear energy.

The race to establish a lasting human presence on the Moon is no longer confined to science fiction or the distant pages of government planning documents. NASA’s Artemis campaign is preparing for a return to the lunar surface, China and Russia are advancing plans for an International Lunar Research Station in the 2030s, and European and commercial partners are developing the landers, habitats, communications systems, and other infrastructure that sustained exploration will require. The schedules will change, as spaceflight schedules invariably do. The larger ambition will not.

Yet beneath the rockets, budgets, and geopolitical theater lies a stubborn engineering question: How can a lunar settlement maintain a continuous supply of electricity through the Moon’s long night?

This is not a peripheral technical problem. It is the foundation on which almost everything else depends. Without reliable power, life-support systems fail, temperatures become lethal, communications go dark, and scientific instruments shut down. On Earth, a blackout is disruptive. On the Moon, where no rescue crew can arrive quickly, it can become a fatal emergency.

Across much of the lunar surface, one full day-night cycle lasts about 29.5 Earth days—roughly two weeks of sunlight followed by two weeks of darkness. During the lunar night, solar panels cease producing electricity, while surface temperatures in some locations can plunge below minus 170 degrees Celsius. Any outpost designed to endure rather than merely visit must survive that darkness without suspending the systems that keep its crew alive.

Apollo did not face this problem on anything like the same scale. Between 1969 and 1972, astronauts landed during lunar daylight, worked for hours or days, and departed before sunset. Batteries and fuel cells supported finite missions with tightly controlled schedules. A settlement is different. Life support, thermal control, communications, research equipment, and eventually industrial machinery must operate continuously. A base cannot simply close for half of every lunar month.

Solar power paired with batteries is the most familiar proposed solution. The concept is straightforward: generate surplus electricity during the lunar day and store it for use at night. The arithmetic, however, quickly becomes punishing. A modest habitat drawing 10 kilowatts continuously would consume approximately 3,540 kilowatt-hours over a 354-hour night. At an optimistic cell-level energy density of 250 watt-hours per kilogram, storing that much electricity would require more than 14 metric tons of batteries before accounting for packaging, thermal management, power electronics, degradation, redundancy, or emergency reserves.

And 10 kilowatts would cover only a small outpost’s most basic needs. It would not sustain extensive mining, oxygen production, scientific laboratories, construction equipment, or fuel manufacturing. As the settlement grew, its storage requirement would grow with it. Batteries would also lose capacity and eventually need replacement. Shipping many tons of new storage hardware from Earth every few years would impose an enormous logistical cost on a project whose long-term viability depends on reducing resupply.

The lunar south pole makes the solar case more attractive, but not decisive. Some elevated ridges receive sunlight for much of the year, which is one reason mission planners favor the region. Yet illumination varies with location, terrain, season, and the height of the solar array. Near-continuous sunlight is not continuous sunlight, and even relatively brief outages require backup generation or substantial storage. The Sun also remains low on the horizon, complicating the placement of arrays and encouraging tall vertical structures that add mass and engineering risk.

Then there is lunar dust. Fine, abrasive, and electrostatically charged, it clung to spacesuits, tools, and equipment during the Apollo missions. On a permanent base, dust deposited on solar arrays could steadily diminish their output. The Moon has no wind or rain to clean exposed surfaces, so the arrays would require active dust mitigation, regular maintenance, or both. Solar power will almost certainly be part of a lunar energy system, but its exposure to darkness, terrain, and dust makes it a fragile foundation for a settlement on its own.

Radioisotope thermoelectric generators, or RTGs, offer a different form of reliability. They convert heat from radioactive decay into electricity and have powered missions including Voyager, Curiosity, and Perseverance. They function without sunlight, contain few moving parts, and can operate for years. Their great weakness is scale.

A modern space RTG produces only about 110 watts of electricity at the beginning of a mission. Supplying even a 10-kilowatt habitat would therefore require roughly 90 units; a more capable 40-kilowatt installation would require several hundred. The necessary fuel, plutonium-238, is scarce, costly, and produced in limited quantities. RTGs are exceptionally useful for spacecraft, instruments, and robotic missions, but they cannot realistically serve as the main power plant for a human settlement.

That leaves nuclear fission as the only credible backbone for a permanent lunar power grid. A compact fission reactor can produce electricity continuously, irrespective of sunlight. It does not have to store two weeks of energy in advance, and it can deliver tens of kilowatts rather than hundreds of watts. Properly designed, it could operate for years with limited intervention while providing the steady output needed for life support, science, mobility, and industry.

The technology is not merely theoretical. In 2018, NASA and the U.S. Department of Energy completed the Kilopower Reactor Using Stirling Technology, or KRUSTY, experiment, demonstrating a small fission system intended for space applications. In 2022, NASA awarded three industry contracts for preliminary designs of a lunar fission surface power system. The specifications called for a reactor producing 40 kilowatts of electricity, weighing less than six metric tons, and supporting operations for at least a decade. In January 2026, NASA and the Energy Department renewed that commitment, announcing an effort to develop a lunar surface reactor for deployment by 2030.

Fission becomes still more compelling when the discussion moves beyond survival to settlement. A durable human presence will depend on in-situ resource utilization: extracting water, oxygen, and fuel from lunar material rather than launching every kilogram from Earth. Transportation costs make permanent dependence on terrestrial resupply economically and operationally untenable.

Scientists have confirmed water ice in permanently shadowed regions near the lunar poles. If it can be mined and processed at useful scale, that ice could yield drinking water, breathable oxygen, and hydrogen and oxygen for rocket propellant. But the relevant machinery—excavators, heaters, purification systems, and electrolysis plants—will demand substantial, dependable electricity, often in places sunlight never reaches.

Producing oxygen from lunar regolith presents a similar challenge. Many proposed methods require material to be heated to extremely high temperatures for extended periods. Repeatedly interrupting those processes during periods of darkness would waste energy, reduce output, and subject equipment to damaging thermal cycles. Industry needs predictable baseload power, not merely an abundant supply when conditions happen to be favorable.

A fission reactor can provide that power independently of local illumination and can be installed at a safe distance from living quarters. Additional units could be added as demand rises, allowing an outpost to grow from a scientific station into a larger industrial settlement. Solar arrays and batteries would still be valuable: they could supplement the reactor, serve distributed equipment, and add redundancy. The strongest architecture is likely to be a diversified microgrid. But diversity does not eliminate the need for a dependable core.

None of this makes lunar nuclear power simple. A reactor must survive launch, landing, abrasive dust, radiation, wide temperature swings, and years without conventional maintenance. Engineers must manage waste heat in a vacuum, bury or distance the reactor to limit crew exposure, and design autonomous safety systems. Launch accidents also demand containment measures that prevent the dispersal of nuclear material.

The political and legal questions are equally consequential. As several countries pursue lunar infrastructure, they will need transparent standards governing reactor siting, operational safety, emergency response, exclusion zones, and end-of-life disposal. A power source essential to lunar settlement could also become a source of mistrust if states deploy it without meaningful notification or consultation. Technical capability must therefore advance alongside rules that reduce the risk of accident and geopolitical confrontation.

Competing technologies may improve. Batteries will become lighter, regenerative fuel cells may help bridge periods of darkness, and power-beaming systems could eventually transmit electricity between illuminated and shadowed locations. Those options deserve investment. But none currently combines the energy density, endurance, scale, and independence from local conditions that fission provides.

The Moon’s long night, permanently shadowed craters, and extreme temperatures are not policy choices that can be negotiated away. Solar arrays can make a major contribution during periods of illumination. Batteries can smooth demand and provide emergency backup. RTGs can power specialized equipment. Each has a role, but none can yet carry the central burden of a growing settlement.

For the foreseeable future, nuclear fission is therefore more than the strongest option. It is the only serious candidate for the continuous, high-density baseload power required to turn a temporary lunar outpost into a permanent human presence. The countries now competing to return to the Moon can postpone that conclusion, but they cannot engineer around it. If humanity intends to stay, it will need a reactor.